diff --git a/ci/run_ctests.sh b/ci/run_ctests.sh
index f6d0ef26ef..3086c39e8c 100755
--- a/ci/run_ctests.sh
+++ b/ci/run_ctests.sh
@@ -6,4 +6,4 @@ set -euo pipefail
 # Support customizing the ctests' install location
 cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libraft/"
 
-ctest --output-on-failure "$@"
+ctest --output-on-failure --no-tests=error "$@"
diff --git a/conda/environments/all_cuda-118_arch-aarch64.yaml b/conda/environments/all_cuda-118_arch-aarch64.yaml
index 613b44dfb7..40b031d677 100644
--- a/conda/environments/all_cuda-118_arch-aarch64.yaml
+++ b/conda/environments/all_cuda-118_arch-aarch64.yaml
@@ -39,7 +39,7 @@ dependencies:
 - nccl>=2.9.9
 - ninja
 - numba>=0.57
-- numpy>=1.21
+- numpy>=1.23
 - numpydoc
 - nvcc_linux-aarch64=11.8
 - pre-commit
diff --git a/conda/environments/all_cuda-118_arch-x86_64.yaml b/conda/environments/all_cuda-118_arch-x86_64.yaml
index 69a7a17fc0..5485d09a37 100644
--- a/conda/environments/all_cuda-118_arch-x86_64.yaml
+++ b/conda/environments/all_cuda-118_arch-x86_64.yaml
@@ -39,7 +39,7 @@ dependencies:
 - nccl>=2.9.9
 - ninja
 - numba>=0.57
-- numpy>=1.21
+- numpy>=1.23
 - numpydoc
 - nvcc_linux-64=11.8
 - pre-commit
diff --git a/conda/environments/all_cuda-122_arch-aarch64.yaml b/conda/environments/all_cuda-122_arch-aarch64.yaml
index 1c924bae46..b688bf3952 100644
--- a/conda/environments/all_cuda-122_arch-aarch64.yaml
+++ b/conda/environments/all_cuda-122_arch-aarch64.yaml
@@ -36,7 +36,7 @@ dependencies:
 - nccl>=2.9.9
 - ninja
 - numba>=0.57
-- numpy>=1.21
+- numpy>=1.23
 - numpydoc
 - pre-commit
 - pydata-sphinx-theme
diff --git a/conda/environments/all_cuda-122_arch-x86_64.yaml b/conda/environments/all_cuda-122_arch-x86_64.yaml
index 09da75016a..013f852aee 100644
--- a/conda/environments/all_cuda-122_arch-x86_64.yaml
+++ b/conda/environments/all_cuda-122_arch-x86_64.yaml
@@ -36,7 +36,7 @@ dependencies:
 - nccl>=2.9.9
 - ninja
 - numba>=0.57
-- numpy>=1.21
+- numpy>=1.23
 - numpydoc
 - pre-commit
 - pydata-sphinx-theme
diff --git a/conda/recipes/pylibraft/meta.yaml b/conda/recipes/pylibraft/meta.yaml
index d2b709172e..5c2829d297 100644
--- a/conda/recipes/pylibraft/meta.yaml
+++ b/conda/recipes/pylibraft/meta.yaml
@@ -65,7 +65,7 @@ requirements:
     {% endif %}
     - libraft {{ version }}
     - libraft-headers {{ version }}
-    - numpy >=1.21
+    - numpy >=1.23
     - python x.x
     - rmm ={{ minor_version }}
 
diff --git a/cpp/.clang-format b/cpp/.clang-format
index 18f376d660..e939dc3c16 100644
--- a/cpp/.clang-format
+++ b/cpp/.clang-format
@@ -27,7 +27,7 @@ AlwaysBreakAfterDefinitionReturnType: None
 AlwaysBreakAfterReturnType: None
 AlwaysBreakBeforeMultilineStrings: true
 AlwaysBreakTemplateDeclarations: Yes
-BinPackArguments:  false       
+BinPackArguments:  false
 BinPackParameters: false
 BraceWrapping:
   AfterClass:            false
@@ -71,8 +71,28 @@ ForEachMacros:
   - foreach
   - Q_FOREACH
   - BOOST_FOREACH
-IncludeBlocks: Preserve
-IncludeIsMainRegex: '([-_](test|unittest))?$'
+IncludeBlocks: Regroup
+IncludeCategories:
+  - Regex:           '^"' # quoted includes
+    Priority:        1
+  - Regex:           '^<(common|benchmarks|tests)/' # benchmark/test includes
+    Priority:        2
+  #- Regex:           '^<(cuspatial_test|cuproj_test)/' # cuSpatial/cuProj test includes
+  #  Priority:        3
+  - Regex:           '^<raft/' # RAFT includes
+    Priority:        3
+  - Regex:           '^<(cudf|cuml|raft|kvikio)' # Other RAPIDS includes
+    Priority:        4
+  - Regex:           '^<rmm/' # RMM includes
+    Priority:        5
+  - Regex:           '^<(thrust|cub|cuda)/' # CCCL includes
+    Priority:        6
+  - Regex:           '^<(cooperative_groups|cuco|cuda|device_types|driver_types|math_constants|nvtx3)' # CUDA includes
+    Priority:        6
+  - Regex:           '^<.*\..*' # other system includes (e.g. with a '.')
+    Priority:        7
+  - Regex:           '^<[^.]+' # STL includes (no '.')
+    Priority:        8
 IndentCaseLabels: true
 IndentPPDirectives: None
 IndentWidth:     2
diff --git a/cpp/bench/ann/src/common/benchmark.cpp b/cpp/bench/ann/src/common/benchmark.cpp
index 6424a36471..5510abf42f 100644
--- a/cpp/bench/ann/src/common/benchmark.cpp
+++ b/cpp/bench/ann/src/common/benchmark.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,19 +13,21 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+
+// clang-format off
 #include "cuda_stub.hpp"  // must go first
+// clang-format on
 
 #include "ann_types.hpp"
 
+#include <dlfcn.h>
 #define JSON_DIAGNOSTICS 1
 #include <nlohmann/json.hpp>
 
+#include <filesystem>
 #include <memory>
 #include <unordered_map>
 
-#include <dlfcn.h>
-#include <filesystem>
-
 namespace raft::bench::ann {
 
 struct lib_handle {
diff --git a/cpp/bench/ann/src/common/benchmark.hpp b/cpp/bench/ann/src/common/benchmark.hpp
index 851993712c..d7bcd17a00 100644
--- a/cpp/bench/ann/src/common/benchmark.hpp
+++ b/cpp/bench/ann/src/common/benchmark.hpp
@@ -20,9 +20,11 @@
 #include "dataset.hpp"
 #include "util.hpp"
 
-#include <benchmark/benchmark.h>
 #include <raft/core/logger.hpp>
 
+#include <benchmark/benchmark.h>
+#include <unistd.h>
+
 #include <algorithm>
 #include <chrono>
 #include <cmath>
@@ -34,7 +36,6 @@
 #include <numeric>
 #include <sstream>
 #include <string>
-#include <unistd.h>
 #include <vector>
 
 namespace raft::bench::ann {
diff --git a/cpp/bench/ann/src/common/cuda_huge_page_resource.hpp b/cpp/bench/ann/src/common/cuda_huge_page_resource.hpp
index ef071c7026..27be26dfe9 100644
--- a/cpp/bench/ann/src/common/cuda_huge_page_resource.hpp
+++ b/cpp/bench/ann/src/common/cuda_huge_page_resource.hpp
@@ -15,12 +15,15 @@
  */
 #pragma once
 
+#include <raft/core/error.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
-#include <rmm/detail/error.hpp>
 #include <rmm/mr/device/device_memory_resource.hpp>
+
 #include <sys/mman.h>
 
 #include <cstddef>
+#include <cstring>
 
 namespace raft::mr {
 /**
diff --git a/cpp/bench/ann/src/common/cuda_pinned_resource.hpp b/cpp/bench/ann/src/common/cuda_pinned_resource.hpp
index a86a5df809..3256fc293c 100644
--- a/cpp/bench/ann/src/common/cuda_pinned_resource.hpp
+++ b/cpp/bench/ann/src/common/cuda_pinned_resource.hpp
@@ -15,10 +15,9 @@
  */
 #pragma once
 
-#include <rmm/mr/device/device_memory_resource.hpp>
-
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/detail/error.hpp>
+#include <rmm/mr/device/device_memory_resource.hpp>
 
 #include <cstddef>
 
diff --git a/cpp/bench/ann/src/common/cuda_stub.hpp b/cpp/bench/ann/src/common/cuda_stub.hpp
index 512ccbce34..5ed138a86d 100644
--- a/cpp/bench/ann/src/common/cuda_stub.hpp
+++ b/cpp/bench/ann/src/common/cuda_stub.hpp
@@ -13,7 +13,6 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-#pragma once
 
 /*
 The content of this header is governed by two preprocessor definitions:
@@ -21,26 +20,30 @@ The content of this header is governed by two preprocessor definitions:
   - BUILD_CPU_ONLY - whether none of the CUDA functions are used.
   - ANN_BENCH_LINK_CUDART - dynamically link against this string if defined.
 
-______________________________________________________________________________
+___________________________________________________________________________________
 |BUILD_CPU_ONLY | ANN_BENCH_LINK_CUDART |         cudart      | cuda_runtime_api.h |
-|         |                       |  found    |  needed |      included      |
-|---------|-----------------------|-----------|---------|--------------------|
-|   ON    |    <not defined>      |  false    |  false  |       NO           |
-|   ON    |   "cudart.so.xx.xx"   |  false    |  false  |       NO           |
-|  OFF    |     <nod defined>     |   true    |   true  |      YES           |
-|  OFF    |   "cudart.so.xx.xx"   | <runtime> |   true  |      YES           |
-------------------------------------------------------------------------------
+|               |                       |  found    |  needed |      included      |
+|---------------|-----------------------|-----------|---------|--------------------|
+|   ON          |    <not defined>      |  false    |  false  |       NO           |
+|   ON          |   "cudart.so.xx.xx"   |  false    |  false  |       NO           |
+|  OFF          |     <not defined>     |   true    |   true  |      YES           |
+|  OFF          |   "cudart.so.xx.xx"   | <runtime> |   true  |      YES           |
+------------------------------------------------------------------------------------
 */
 
+#pragma once
+
 #ifndef BUILD_CPU_ONLY
 #include <cuda_fp16.h>
 #include <cuda_runtime_api.h>
 #ifdef ANN_BENCH_LINK_CUDART
-#include <cstring>
 #include <dlfcn.h>
+
+#include <cstring>
 #endif
 #else
 #include <cstdint>
+
 typedef void* cudaStream_t;
 typedef void* cudaEvent_t;
 typedef uint16_t half;
diff --git a/cpp/bench/ann/src/common/thread_pool.hpp b/cpp/bench/ann/src/common/thread_pool.hpp
index c01fa2c32c..4a5684ecb3 100644
--- a/cpp/bench/ann/src/common/thread_pool.hpp
+++ b/cpp/bench/ann/src/common/thread_pool.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
+#include <omp.h>
+
 #include <atomic>
 #include <future>
 #include <memory>
 #include <mutex>
-#include <omp.h>
 #include <stdexcept>
 #include <thread>
 #include <utility>
diff --git a/cpp/bench/ann/src/common/util.hpp b/cpp/bench/ann/src/common/util.hpp
index 8f6468befa..6cdff316e9 100644
--- a/cpp/bench/ann/src/common/util.hpp
+++ b/cpp/bench/ann/src/common/util.hpp
@@ -25,21 +25,20 @@
 #include <sys/stat.h>
 #include <sys/types.h>
 
+#include <atomic>
 #include <chrono>
+#include <condition_variable>
 #include <cstdio>
 #include <cstring>
 #include <ctime>
+#include <filesystem>
+#include <functional>
 #include <mutex>
+#include <optional>
 #include <sstream>
 #include <string>
 #include <vector>
 
-#include <atomic>
-#include <condition_variable>
-#include <filesystem>
-#include <functional>
-#include <optional>
-
 namespace raft::bench::ann {
 
 /**
diff --git a/cpp/bench/ann/src/faiss/faiss_cpu_benchmark.cpp b/cpp/bench/ann/src/faiss/faiss_cpu_benchmark.cpp
index e3e25a99a2..965522d929 100644
--- a/cpp/bench/ann/src/faiss/faiss_cpu_benchmark.cpp
+++ b/cpp/bench/ann/src/faiss/faiss_cpu_benchmark.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,6 +14,12 @@
  * limitations under the License.
  */
 
+#include "../common/ann_types.hpp"
+#include "faiss_cpu_wrapper.h"
+
+#define JSON_DIAGNOSTICS 1
+#include <nlohmann/json.hpp>
+
 #include <algorithm>
 #include <cmath>
 #include <memory>
@@ -22,11 +28,6 @@
 #include <type_traits>
 #include <utility>
 
-#include "../common/ann_types.hpp"
-#include "faiss_cpu_wrapper.h"
-#define JSON_DIAGNOSTICS 1
-#include <nlohmann/json.hpp>
-
 namespace raft::bench::ann {
 
 template <typename T>
diff --git a/cpp/bench/ann/src/faiss/faiss_gpu_benchmark.cu b/cpp/bench/ann/src/faiss/faiss_gpu_benchmark.cu
index a9388531cc..c5056cb364 100644
--- a/cpp/bench/ann/src/faiss/faiss_gpu_benchmark.cu
+++ b/cpp/bench/ann/src/faiss/faiss_gpu_benchmark.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,6 +14,14 @@
  * limitations under the License.
  */
 
+#include "../common/ann_types.hpp"
+
+#undef WARP_SIZE
+#include "faiss_gpu_wrapper.h"
+
+#define JSON_DIAGNOSTICS 1
+#include <nlohmann/json.hpp>
+
 #include <algorithm>
 #include <cmath>
 #include <memory>
@@ -22,12 +30,6 @@
 #include <type_traits>
 #include <utility>
 
-#include "../common/ann_types.hpp"
-#undef WARP_SIZE
-#include "faiss_gpu_wrapper.h"
-#define JSON_DIAGNOSTICS 1
-#include <nlohmann/json.hpp>
-
 namespace raft::bench::ann {
 
 template <typename T>
diff --git a/cpp/bench/ann/src/ggnn/ggnn_benchmark.cu b/cpp/bench/ann/src/ggnn/ggnn_benchmark.cu
index 884476ea12..48d41388d4 100644
--- a/cpp/bench/ann/src/ggnn/ggnn_benchmark.cu
+++ b/cpp/bench/ann/src/ggnn/ggnn_benchmark.cu
@@ -14,6 +14,12 @@
  * limitations under the License.
  */
 
+#include "../common/ann_types.hpp"
+#include "ggnn_wrapper.cuh"
+
+#define JSON_DIAGNOSTICS 1
+#include <nlohmann/json.hpp>
+
 #include <algorithm>
 #include <cmath>
 #include <memory>
@@ -22,11 +28,6 @@
 #include <type_traits>
 #include <utility>
 
-#include "../common/ann_types.hpp"
-#include "ggnn_wrapper.cuh"
-#define JSON_DIAGNOSTICS 1
-#include <nlohmann/json.hpp>
-
 namespace raft::bench::ann {
 
 template <typename T>
diff --git a/cpp/bench/ann/src/ggnn/ggnn_wrapper.cuh b/cpp/bench/ann/src/ggnn/ggnn_wrapper.cuh
index 5f5013ef7f..c89f02d974 100644
--- a/cpp/bench/ann/src/ggnn/ggnn_wrapper.cuh
+++ b/cpp/bench/ann/src/ggnn/ggnn_wrapper.cuh
@@ -19,9 +19,10 @@
 #include "../common/ann_types.hpp"
 #include "../common/util.hpp"
 
-#include <ggnn/cuda_knn_ggnn_gpu_instance.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <ggnn/cuda_knn_ggnn_gpu_instance.cuh>
+
 #include <memory>
 #include <stdexcept>
 
diff --git a/cpp/bench/ann/src/hnswlib/hnswlib_benchmark.cpp b/cpp/bench/ann/src/hnswlib/hnswlib_benchmark.cpp
index 1af19a22cb..df82c68830 100644
--- a/cpp/bench/ann/src/hnswlib/hnswlib_benchmark.cpp
+++ b/cpp/bench/ann/src/hnswlib/hnswlib_benchmark.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,10 @@
  */
 
 #include "../common/ann_types.hpp"
+#include "hnswlib_wrapper.h"
+
+#define JSON_DIAGNOSTICS 1
+#include <nlohmann/json.hpp>
 
 #include <algorithm>
 #include <cmath>
@@ -24,10 +28,6 @@
 #include <type_traits>
 #include <utility>
 
-#include "hnswlib_wrapper.h"
-#define JSON_DIAGNOSTICS 1
-#include <nlohmann/json.hpp>
-
 namespace raft::bench::ann {
 
 template <typename T>
diff --git a/cpp/bench/ann/src/hnswlib/hnswlib_wrapper.h b/cpp/bench/ann/src/hnswlib/hnswlib_wrapper.h
index 45f7698a2d..a8f7dd824f 100644
--- a/cpp/bench/ann/src/hnswlib/hnswlib_wrapper.h
+++ b/cpp/bench/ann/src/hnswlib/hnswlib_wrapper.h
@@ -15,6 +15,11 @@
  */
 #pragma once
 
+#include "../common/ann_types.hpp"
+#include "../common/thread_pool.hpp"
+
+#include <hnswlib/hnswlib.h>
+
 #include <algorithm>
 #include <atomic>
 #include <cassert>
@@ -31,10 +36,6 @@
 #include <utility>
 #include <vector>
 
-#include "../common/ann_types.hpp"
-#include "../common/thread_pool.hpp"
-#include <hnswlib/hnswlib.h>
-
 namespace raft::bench::ann {
 
 template <typename T>
diff --git a/cpp/bench/ann/src/raft/raft_cagra_hnswlib_wrapper.h b/cpp/bench/ann/src/raft/raft_cagra_hnswlib_wrapper.h
index 2018815c1e..ed9c120ed4 100644
--- a/cpp/bench/ann/src/raft/raft_cagra_hnswlib_wrapper.h
+++ b/cpp/bench/ann/src/raft/raft_cagra_hnswlib_wrapper.h
@@ -17,6 +17,7 @@
 
 #include "../hnswlib/hnswlib_wrapper.h"
 #include "raft_cagra_wrapper.h"
+
 #include <memory>
 
 namespace raft::bench::ann {
diff --git a/cpp/bench/ann/src/raft/raft_cagra_wrapper.h b/cpp/bench/ann/src/raft/raft_cagra_wrapper.h
old mode 100755
new mode 100644
index ba4936976e..25f7f93777
--- a/cpp/bench/ann/src/raft/raft_cagra_wrapper.h
+++ b/cpp/bench/ann/src/raft/raft_cagra_wrapper.h
@@ -15,11 +15,11 @@
  */
 #pragma once
 
-#include <cassert>
-#include <fstream>
-#include <iostream>
-#include <memory>
-#include <optional>
+#include "../common/ann_types.hpp"
+#include "../common/cuda_huge_page_resource.hpp"
+#include "../common/cuda_pinned_resource.hpp"
+#include "raft_ann_bench_utils.h"
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/logger.hpp>
@@ -33,21 +33,19 @@
 #include <raft/neighbors/ivf_pq_types.hpp>
 #include <raft/neighbors/nn_descent_types.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
+#include <rmm/mr/device/device_memory_resource.hpp>
+
+#include <cassert>
+#include <fstream>
+#include <iostream>
+#include <memory>
+#include <optional>
 #include <stdexcept>
 #include <string>
 #include <type_traits>
 
-#include "../common/ann_types.hpp"
-#include "raft_ann_bench_utils.h"
-#include <raft/util/cudart_utils.hpp>
-
-#include "../common/cuda_huge_page_resource.hpp"
-#include "../common/cuda_pinned_resource.hpp"
-
-#include <rmm/device_uvector.hpp>
-#include <rmm/mr/device/device_memory_resource.hpp>
-
 namespace raft::bench::ann {
 
 enum class AllocatorType { HostPinned, HostHugePage, Device };
diff --git a/cpp/bench/ann/src/raft/raft_wrapper.h b/cpp/bench/ann/src/raft/raft_wrapper.h
index 6fcb869d38..586b81ae06 100644
--- a/cpp/bench/ann/src/raft/raft_wrapper.h
+++ b/cpp/bench/ann/src/raft/raft_wrapper.h
@@ -15,20 +15,21 @@
  */
 #pragma once
 
-#include <cassert>
-#include <memory>
+#include "../common/ann_types.hpp"
+#include "raft_ann_bench_utils.h"
+
 #include <raft/core/device_resources.hpp>
 #include <raft/distance/detail/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/brute_force.cuh>
 #include <raft/neighbors/brute_force_serialize.cuh>
+
+#include <cassert>
+#include <memory>
 #include <stdexcept>
 #include <string>
 #include <type_traits>
 
-#include "../common/ann_types.hpp"
-#include "raft_ann_bench_utils.h"
-
 namespace raft_temp {
 
 inline raft::distance::DistanceType parse_metric_type(raft::bench::ann::Metric metric)
diff --git a/cpp/bench/prims/cluster/kmeans.cu b/cpp/bench/prims/cluster/kmeans.cu
index 3147960f72..6387211135 100644
--- a/cpp/bench/prims/cluster/kmeans.cu
+++ b/cpp/bench/prims/cluster/kmeans.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/cluster/kmeans.cuh>
 #include <raft/cluster/kmeans_types.hpp>
 
diff --git a/cpp/bench/prims/cluster/kmeans_balanced.cu b/cpp/bench/prims/cluster/kmeans_balanced.cu
index 129578c303..dc05783989 100644
--- a/cpp/bench/prims/cluster/kmeans_balanced.cu
+++ b/cpp/bench/prims/cluster/kmeans_balanced.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/cluster/kmeans_balanced.cuh>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
diff --git a/cpp/bench/prims/common/benchmark.hpp b/cpp/bench/prims/common/benchmark.hpp
index c924d8b50a..4ecad6df3d 100644
--- a/cpp/bench/prims/common/benchmark.hpp
+++ b/cpp/bench/prims/common/benchmark.hpp
@@ -16,18 +16,14 @@
 
 #pragma once
 
-#include <memory>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/interruptible.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/make_blobs.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <benchmark/benchmark.h>
-
 #include <rmm/cuda_device.hpp>
 #include <rmm/cuda_stream.hpp>
 #include <rmm/cuda_stream_view.hpp>
@@ -35,6 +31,10 @@
 #include <rmm/mr/device/per_device_resource.hpp>
 #include <rmm/mr/device/pool_memory_resource.hpp>
 
+#include <benchmark/benchmark.h>
+
+#include <memory>
+
 namespace raft::bench {
 
 /**
diff --git a/cpp/bench/prims/core/bitset.cu b/cpp/bench/prims/core/bitset.cu
index ce3136bcd5..166f7aa6dc 100644
--- a/cpp/bench/prims/core/bitset.cu
+++ b/cpp/bench/prims/core/bitset.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,10 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/core/bitset.cuh>
 #include <raft/core/device_mdspan.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::core {
diff --git a/cpp/bench/prims/core/copy.cu b/cpp/bench/prims/core/copy.cu
index 31ee83b924..4898f560cb 100644
--- a/cpp/bench/prims/core/copy.cu
+++ b/cpp/bench/prims/core/copy.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
-#include <cstdint>
+
 #include <raft/core/copy.cuh>
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdarray.hpp>
@@ -25,6 +25,8 @@
 #include <raft/core/memory_type.hpp>
 #include <raft/thirdparty/mdspan/include/experimental/mdspan>
 
+#include <cstdint>
+
 namespace raft::bench::core {
 
 template <typename IdxT, std::size_t Rank>
diff --git a/cpp/bench/prims/distance/distance_common.cuh b/cpp/bench/prims/distance/distance_common.cuh
index dff3401b62..8368062168 100644
--- a/cpp/bench/prims/distance/distance_common.cuh
+++ b/cpp/bench/prims/distance/distance_common.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,10 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/distance/distance.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::distance {
diff --git a/cpp/bench/prims/distance/fused_l2_nn.cu b/cpp/bench/prims/distance/fused_l2_nn.cu
index c0ebd60458..a263bef6ba 100644
--- a/cpp/bench/prims/distance/fused_l2_nn.cu
+++ b/cpp/bench/prims/distance/fused_l2_nn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,10 +15,12 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/fused_l2_nn.cuh>
 #include <raft/linalg/norm.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::distance {
diff --git a/cpp/bench/prims/distance/kernels.cu b/cpp/bench/prims/distance/kernels.cu
index 3f74759665..eb86330637 100644
--- a/cpp/bench/prims/distance/kernels.cu
+++ b/cpp/bench/prims/distance/kernels.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,12 +14,14 @@
  * limitations under the License.
  */
 #include <common/benchmark.hpp>
-#include <memory>
+
 #include <raft/core/device_resources.hpp>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/kernels.cuh>
 #include <raft/random/rng.cuh>
+
+#include <memory>
 #include <sstream>
 #include <string>
 #include <vector>
diff --git a/cpp/bench/prims/distance/masked_nn.cu b/cpp/bench/prims/distance/masked_nn.cu
index 19d78f4cd9..979d438b67 100644
--- a/cpp/bench/prims/distance/masked_nn.cu
+++ b/cpp/bench/prims/distance/masked_nn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,14 +14,8 @@
  * limitations under the License.
  */
 
-#include <cstdint>
-#include <fstream>
-#include <iostream>
-#include <sstream>
-#include <string>
-
 #include <common/benchmark.hpp>
-#include <limits>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/handle.hpp>
@@ -30,6 +24,13 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cstdint>
+#include <fstream>
+#include <iostream>
+#include <limits>
+#include <sstream>
+#include <string>
+
 namespace raft::bench::distance::masked_nn {
 
 // Introduce various sparsity patterns
diff --git a/cpp/bench/prims/distance/tune_pairwise/bench.cu b/cpp/bench/prims/distance/tune_pairwise/bench.cu
index 87159ab1b1..81105cdefe 100644
--- a/cpp/bench/prims/distance/tune_pairwise/bench.cu
+++ b/cpp/bench/prims/distance/tune_pairwise/bench.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -38,12 +38,16 @@
 // maximal throughput of a kernel. Measuring other things, like performance on
 // skinny or wide matrices is not yet implemented.
 
-#include "kernel.cuh"                                       // launch_kernel
-#include <algorithm>                                        // std::min
-#include <common/benchmark.hpp>                             // RAFT_BENCH_REGISTER
+#include "kernel.cuh"  // launch_kernel
+
+#include <common/benchmark.hpp>  // RAFT_BENCH_REGISTER
+
 #include <raft/distance/detail/pairwise_matrix/params.cuh>  // pairwise_matrix_params
-#include <rmm/device_uvector.hpp>                           // rmm::device_uvector
-#include <vector>                                           // std::vector
+
+#include <rmm/device_uvector.hpp>  // rmm::device_uvector
+
+#include <algorithm>  // std::min
+#include <vector>     // std::vector
 
 namespace raft::bench::distance::tune {
 
diff --git a/cpp/bench/prims/distance/tune_pairwise/kernel.cu b/cpp/bench/prims/distance/tune_pairwise/kernel.cu
index 3112e1ea9a..42173c51f5 100644
--- a/cpp/bench/prims/distance/tune_pairwise/kernel.cu
+++ b/cpp/bench/prims/distance/tune_pairwise/kernel.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "kernel.cuh"
+
 #include <raft/distance/detail/pairwise_matrix/kernel_sm60.cuh>  // pairwise_matrix_sm60_wrapper
 #include <raft/linalg/contractions.cuh>                          // raft::linalg::Policy4x4
 #include <raft/util/arch.cuh>  // raft::util::arch::SM_compute_arch
diff --git a/cpp/bench/prims/linalg/add.cu b/cpp/bench/prims/linalg/add.cu
index 456214ad7b..f861e24d44 100644
--- a/cpp/bench/prims/linalg/add.cu
+++ b/cpp/bench/prims/linalg/add.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,9 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/add.cuh>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::linalg {
diff --git a/cpp/bench/prims/linalg/map_then_reduce.cu b/cpp/bench/prims/linalg/map_then_reduce.cu
index 84aebd85bf..3ca775a1da 100644
--- a/cpp/bench/prims/linalg/map_then_reduce.cu
+++ b/cpp/bench/prims/linalg/map_then_reduce.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,9 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/map_then_reduce.cuh>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::linalg {
diff --git a/cpp/bench/prims/linalg/matrix_vector_op.cu b/cpp/bench/prims/linalg/matrix_vector_op.cu
index d1fbaee79b..c95efed91b 100644
--- a/cpp/bench/prims/linalg/matrix_vector_op.cu
+++ b/cpp/bench/prims/linalg/matrix_vector_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,9 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/matrix_vector_op.cuh>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft::bench::linalg {
diff --git a/cpp/bench/prims/linalg/norm.cu b/cpp/bench/prims/linalg/norm.cu
index 1db23e4ca4..7d8a07e932 100644
--- a/cpp/bench/prims/linalg/norm.cu
+++ b/cpp/bench/prims/linalg/norm.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/matrix_vector_op.cuh>
 #include <raft/linalg/norm.cuh>
 #include <raft/random/rng.cuh>
diff --git a/cpp/bench/prims/linalg/normalize.cu b/cpp/bench/prims/linalg/normalize.cu
index 91319e774c..7d2a725914 100644
--- a/cpp/bench/prims/linalg/normalize.cu
+++ b/cpp/bench/prims/linalg/normalize.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/normalize.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/itertools.hpp>
diff --git a/cpp/bench/prims/linalg/reduce.cu b/cpp/bench/prims/linalg/reduce.cu
index cf41c5916a..f0fd13529d 100644
--- a/cpp/bench/prims/linalg/reduce.cu
+++ b/cpp/bench/prims/linalg/reduce.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/reduce.cuh>
 
 #include <rmm/device_uvector.hpp>
diff --git a/cpp/bench/prims/linalg/reduce_cols_by_key.cu b/cpp/bench/prims/linalg/reduce_cols_by_key.cu
index 1b584e80c8..f936a04ba1 100644
--- a/cpp/bench/prims/linalg/reduce_cols_by_key.cu
+++ b/cpp/bench/prims/linalg/reduce_cols_by_key.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/reduce_cols_by_key.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/itertools.hpp>
diff --git a/cpp/bench/prims/linalg/reduce_rows_by_key.cu b/cpp/bench/prims/linalg/reduce_rows_by_key.cu
index b68cefc274..a63b679bb6 100644
--- a/cpp/bench/prims/linalg/reduce_rows_by_key.cu
+++ b/cpp/bench/prims/linalg/reduce_rows_by_key.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/linalg/reduce_rows_by_key.cuh>
 #include <raft/random/rng.cuh>
 
diff --git a/cpp/bench/prims/linalg/sddmm.cu b/cpp/bench/prims/linalg/sddmm.cu
index 139a2b838d..cd4259106d 100644
--- a/cpp/bench/prims/linalg/sddmm.cu
+++ b/cpp/bench/prims/linalg/sddmm.cu
@@ -14,17 +14,18 @@
  * limitations under the License.
  */
 #include <common/benchmark.hpp>
-#include <cusparse_v2.h>
+
 #include <raft/core/device_resources.hpp>
 #include <raft/core/resource/cublas_handle.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/sparse/linalg/sddmm.hpp>
 #include <raft/util/itertools.hpp>
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resources.hpp>
+#include <cusparse_v2.h>
 
 #include <random>
 #include <sstream>
diff --git a/cpp/bench/prims/matrix/argmin.cu b/cpp/bench/prims/matrix/argmin.cu
index afee81aa00..29655998e1 100644
--- a/cpp/bench/prims/matrix/argmin.cu
+++ b/cpp/bench/prims/matrix/argmin.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/argmin.cuh>
 #include <raft/random/rng.cuh>
diff --git a/cpp/bench/prims/matrix/gather.cu b/cpp/bench/prims/matrix/gather.cu
index 00a145ffa9..e6f26ba925 100644
--- a/cpp/bench/prims/matrix/gather.cu
+++ b/cpp/bench/prims/matrix/gather.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/gather.cuh>
 #include <raft/random/rng.cuh>
diff --git a/cpp/bench/prims/matrix/main.cpp b/cpp/bench/prims/matrix/main.cpp
index 9cdb1c2546..b59e152971 100644
--- a/cpp/bench/prims/matrix/main.cpp
+++ b/cpp/bench/prims/matrix/main.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <benchmark/benchmark.h>
+
 #include <cstring>
 
 namespace raft::matrix {
diff --git a/cpp/bench/prims/matrix/select_k.cu b/cpp/bench/prims/matrix/select_k.cu
index a8e65d3e03..ff04e6f8a8 100644
--- a/cpp/bench/prims/matrix/select_k.cu
+++ b/cpp/bench/prims/matrix/select_k.cu
@@ -14,19 +14,18 @@
  * limitations under the License.
  */
 
-#include <raft_internal/matrix/select_k.cuh>
-
 #include <common/benchmark.hpp>
 
 #include <raft/core/device_resources.hpp>
 #include <raft/core/nvtx.hpp>
+#include <raft/matrix/detail/select_radix.cuh>
+#include <raft/matrix/detail/select_warpsort.cuh>
+#include <raft/matrix/select_k.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/sparse/detail/utils.h>
 #include <raft/util/cudart_utils.hpp>
 
-#include <raft/matrix/detail/select_radix.cuh>
-#include <raft/matrix/detail/select_warpsort.cuh>
-#include <raft/matrix/select_k.cuh>
+#include <raft_internal/matrix/select_k.cuh>
 
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/per_device_resource.hpp>
diff --git a/cpp/bench/prims/neighbors/cagra_bench.cuh b/cpp/bench/prims/neighbors/cagra_bench.cuh
index 07e93a3473..acbeba375a 100644
--- a/cpp/bench/prims/neighbors/cagra_bench.cuh
+++ b/cpp/bench/prims/neighbors/cagra_bench.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,10 +17,12 @@
 #pragma once
 
 #include <common/benchmark.hpp>
+
 #include <raft/neighbors/cagra.cuh>
 #include <raft/neighbors/sample_filter.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/itertools.hpp>
+
 #include <thrust/sequence.h>
 
 #include <optional>
diff --git a/cpp/bench/prims/neighbors/knn.cuh b/cpp/bench/prims/neighbors/knn.cuh
index 55865b577a..aea7168142 100644
--- a/cpp/bench/prims/neighbors/knn.cuh
+++ b/cpp/bench/prims/neighbors/knn.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,20 +17,18 @@
 #pragma once
 
 #include <common/benchmark.hpp>
-#include <raft/core/resource/device_id.hpp>
-
-#include <raft/random/rng.cuh>
 
 #include <raft/core/bitset.cuh>
+#include <raft/core/resource/device_id.hpp>
 #include <raft/neighbors/ivf_flat.cuh>
 #include <raft/neighbors/ivf_pq.cuh>
 #include <raft/neighbors/sample_filter.cuh>
+#include <raft/random/rng.cuh>
 #include <raft/spatial/knn/knn.cuh>
 #include <raft/util/itertools.hpp>
 
 #include <rmm/mr/device/managed_memory_resource.hpp>
 #include <rmm/mr/device/per_device_resource.hpp>
-
 #include <rmm/mr/host/new_delete_resource.hpp>
 #include <rmm/mr/host/pinned_memory_resource.hpp>
 
diff --git a/cpp/bench/prims/neighbors/refine.cuh b/cpp/bench/prims/neighbors/refine.cuh
index 7cda59b57e..0360babd82 100644
--- a/cpp/bench/prims/neighbors/refine.cuh
+++ b/cpp/bench/prims/neighbors/refine.cuh
@@ -16,8 +16,6 @@
 
 #pragma once
 
-#include <raft_internal/neighbors/refine_helper.cuh>
-
 #include <common/benchmark.hpp>
 
 #include <raft/core/device_mdspan.hpp>
@@ -27,6 +25,8 @@
 #include <raft/neighbors/refine.cuh>
 #include <raft/random/rng.cuh>
 
+#include <raft_internal/neighbors/refine_helper.cuh>
+
 #include <rmm/cuda_device.hpp>
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/mr/device/per_device_resource.hpp>
diff --git a/cpp/bench/prims/neighbors/refine_float_int64_t.cu b/cpp/bench/prims/neighbors/refine_float_int64_t.cu
index bbedc1ae64..d69a157eca 100644
--- a/cpp/bench/prims/neighbors/refine_float_int64_t.cu
+++ b/cpp/bench/prims/neighbors/refine_float_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "refine.cuh"
+
 #include <common/benchmark.hpp>
 
 using namespace raft::neighbors;
diff --git a/cpp/bench/prims/neighbors/refine_uint8_t_int64_t.cu b/cpp/bench/prims/neighbors/refine_uint8_t_int64_t.cu
index 4952361f03..9da536b6c7 100644
--- a/cpp/bench/prims/neighbors/refine_uint8_t_int64_t.cu
+++ b/cpp/bench/prims/neighbors/refine_uint8_t_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "refine.cuh"
+
 #include <common/benchmark.hpp>
 
 using namespace raft::neighbors;
diff --git a/cpp/bench/prims/random/make_blobs.cu b/cpp/bench/prims/random/make_blobs.cu
index f43d914cf2..75ff861926 100644
--- a/cpp/bench/prims/random/make_blobs.cu
+++ b/cpp/bench/prims/random/make_blobs.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,11 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/random/make_blobs.cuh>
+
 #include <rmm/device_uvector.hpp>
+
 #include <vector>
 
 namespace raft::bench::random {
diff --git a/cpp/bench/prims/random/permute.cu b/cpp/bench/prims/random/permute.cu
index 829cf42720..eb6ebc3d73 100644
--- a/cpp/bench/prims/random/permute.cu
+++ b/cpp/bench/prims/random/permute.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/random/permute.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
diff --git a/cpp/bench/prims/random/rng.cu b/cpp/bench/prims/random/rng.cu
index d15c9441d7..037c69e0ea 100644
--- a/cpp/bench/prims/random/rng.cu
+++ b/cpp/bench/prims/random/rng.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <common/benchmark.hpp>
+
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
diff --git a/cpp/bench/prims/sparse/convert_csr.cu b/cpp/bench/prims/sparse/convert_csr.cu
index 634c749a54..687e7850b5 100644
--- a/cpp/bench/prims/sparse/convert_csr.cu
+++ b/cpp/bench/prims/sparse/convert_csr.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,12 +14,14 @@
  * limitations under the License.
  */
 
-#include <stdio.h>
-
 #include <common/benchmark.hpp>
+
 #include <raft/sparse/convert/csr.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <stdio.h>
+
 namespace raft::bench::sparse {
 
 template <typename index_t>
diff --git a/cpp/include/raft/cluster/detail/connectivities.cuh b/cpp/include/raft/cluster/detail/connectivities.cuh
index 49ac6ae704..c527b754c3 100644
--- a/cpp/include/raft/cluster/detail/connectivities.cuh
+++ b/cpp/include/raft/cluster/detail/connectivities.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,21 +16,20 @@
 
 #pragma once
 
+#include <raft/cluster/single_linkage_types.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-
-#include <raft/linalg/unary_op.cuh>
-#include <rmm/device_uvector.hpp>
-
-#include <raft/cluster/single_linkage_types.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
+#include <raft/linalg/unary_op.cuh>
 #include <raft/sparse/convert/csr.cuh>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/neighbors/knn_graph.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+
+#include <rmm/device_uvector.hpp>
 
 #include <thrust/iterator/zip_iterator.h>
 #include <thrust/transform.h>
diff --git a/cpp/include/raft/cluster/detail/kmeans.cuh b/cpp/include/raft/cluster/detail/kmeans.cuh
index e647e33734..4efeedcbaa 100644
--- a/cpp/include/raft/cluster/detail/kmeans.cuh
+++ b/cpp/include/raft/cluster/detail/kmeans.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,19 +15,6 @@
  */
 #pragma once
 
-#include <algorithm>
-#include <cmath>
-#include <cstdio>
-#include <ctime>
-#include <optional>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-#include <random>
-
-#include <cuda.h>
-#include <thrust/fill.h>
-#include <thrust/transform.h>
-
 #include <raft/cluster/detail/kmeans_common.cuh>
 #include <raft/cluster/kmeans_types.hpp>
 #include <raft/common/nvtx.hpp>
@@ -38,6 +25,8 @@
 #include <raft/core/logger.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/map_then_reduce.cuh>
@@ -48,9 +37,21 @@
 #include <raft/matrix/gather.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cuda.h>
+#include <thrust/fill.h>
+#include <thrust/transform.h>
+
+#include <algorithm>
+#include <cmath>
+#include <cstdio>
+#include <ctime>
+#include <optional>
+#include <random>
+
 namespace raft {
 namespace cluster {
 namespace detail {
diff --git a/cpp/include/raft/cluster/detail/kmeans_auto_find_k.cuh b/cpp/include/raft/cluster/detail/kmeans_auto_find_k.cuh
index f6bdb191cd..97755351c4 100644
--- a/cpp/include/raft/cluster/detail/kmeans_auto_find_k.cuh
+++ b/cpp/include/raft/cluster/detail/kmeans_auto_find_k.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,21 +14,18 @@
  * limitations under the License.
  */
 
+#include <raft/cluster/detail/kmeans.cuh>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <thrust/host_vector.h>
-
 #include <raft/core/logger.hpp>
-
-#include <raft/cluster/detail/kmeans.cuh>
-
-#include <raft/core/error.hpp>
-
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/stats/dispersion.cuh>
 
+#include <thrust/host_vector.h>
+
 namespace raft::cluster::detail {
 
 template <typename value_t, typename idx_t>
diff --git a/cpp/include/raft/cluster/detail/kmeans_balanced.cuh b/cpp/include/raft/cluster/detail/kmeans_balanced.cuh
index 3b5d3ff02a..6d3f430e88 100644
--- a/cpp/include/raft/cluster/detail/kmeans_balanced.cuh
+++ b/cpp/include/raft/cluster/detail/kmeans_balanced.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,18 +16,15 @@
 
 #pragma once
 
-#include <limits>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_memory_resource.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-#include <type_traits>
-
 #include <raft/cluster/detail/kmeans_common.cuh>
 #include <raft/cluster/kmeans_balanced_types.hpp>
 #include <raft/common/nvtx.hpp>
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/logger.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/fused_l2_nn.cuh>
@@ -44,7 +41,6 @@
 #include <raft/util/device_atomics.cuh>
 #include <raft/util/integer_utils.hpp>
 
-#include <raft/core/resource/device_memory_resource.hpp>
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_vector.hpp>
@@ -55,7 +51,9 @@
 #include <thrust/gather.h>
 #include <thrust/transform.h>
 
+#include <limits>
 #include <tuple>
+#include <type_traits>
 
 namespace raft::cluster::detail {
 
diff --git a/cpp/include/raft/cluster/detail/kmeans_common.cuh b/cpp/include/raft/cluster/detail/kmeans_common.cuh
index 5d56a1d081..8263aa4615 100644
--- a/cpp/include/raft/cluster/detail/kmeans_common.cuh
+++ b/cpp/include/raft/cluster/detail/kmeans_common.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,20 +15,6 @@
  */
 #pragma once
 
-#include <algorithm>
-#include <cmath>
-#include <cstdio>
-#include <ctime>
-#include <optional>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-#include <random>
-
-#include <cub/cub.cuh>
-#include <cuda.h>
-#include <thrust/fill.h>
-#include <thrust/for_each.h>
-
 #include <raft/cluster/kmeans_types.hpp>
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdarray.hpp>
@@ -36,6 +22,8 @@
 #include <raft/core/logger.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
@@ -51,6 +39,18 @@
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+#include <cuda.h>
+#include <thrust/fill.h>
+#include <thrust/for_each.h>
+
+#include <algorithm>
+#include <cmath>
+#include <cstdio>
+#include <ctime>
+#include <optional>
+#include <random>
+
 namespace raft {
 namespace cluster {
 namespace detail {
diff --git a/cpp/include/raft/cluster/detail/kmeans_deprecated.cuh b/cpp/include/raft/cluster/detail/kmeans_deprecated.cuh
index 0b5dec4e19..e89f5480e3 100644
--- a/cpp/include/raft/cluster/detail/kmeans_deprecated.cuh
+++ b/cpp/include/raft/cluster/detail/kmeans_deprecated.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,13 +21,16 @@
 
 #pragma once
 
-#include <algorithm>
-#include <cmath>
-#include <cstdio>
-#include <ctime>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
+#include <raft/spectral/detail/warn_dbg.hpp>
+#include <raft/spectral/matrix_wrappers.hpp>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+#include <raft/util/device_atomics.cuh>
 
 #include <cuda.h>
 #include <thrust/binary_search.h>
@@ -45,13 +48,10 @@
 #include <thrust/sort.h>
 #include <thrust/transform.h>
 
-#include <raft/core/resources.hpp>
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-#include <raft/spectral/detail/warn_dbg.hpp>
-#include <raft/spectral/matrix_wrappers.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-#include <raft/util/device_atomics.cuh>
+#include <algorithm>
+#include <cmath>
+#include <cstdio>
+#include <ctime>
 
 namespace raft {
 namespace cluster {
diff --git a/cpp/include/raft/cluster/detail/mst.cuh b/cpp/include/raft/cluster/detail/mst.cuh
index a962d4b7c6..55becc8e15 100644
--- a/cpp/include/raft/cluster/detail/mst.cuh
+++ b/cpp/include/raft/cluster/detail/mst.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,12 +17,12 @@
 #pragma once
 
 #include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-
 #include <raft/sparse/neighbors/cross_component_nn.cuh>
 #include <raft/sparse/op/sort.cuh>
 #include <raft/sparse/solver/mst.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/device_ptr.h>
diff --git a/cpp/include/raft/cluster/detail/single_linkage.cuh b/cpp/include/raft/cluster/detail/single_linkage.cuh
index 848ca0357e..ccc6472684 100644
--- a/cpp/include/raft/cluster/detail/single_linkage.cuh
+++ b/cpp/include/raft/cluster/detail/single_linkage.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,14 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cudart_utils.hpp>
-#include <rmm/device_uvector.hpp>
-
 #include <raft/cluster/detail/agglomerative.cuh>
 #include <raft/cluster/detail/connectivities.cuh>
 #include <raft/cluster/detail/mst.cuh>
 #include <raft/cluster/single_linkage_types.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/util/cudart_utils.hpp>
+
+#include <rmm/device_uvector.hpp>
 
 namespace raft::cluster::detail {
 
diff --git a/cpp/include/raft/cluster/kmeans.cuh b/cpp/include/raft/cluster/kmeans.cuh
index d63413e82e..eb28cc1626 100644
--- a/cpp/include/raft/cluster/kmeans.cuh
+++ b/cpp/include/raft/cluster/kmeans.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,6 @@
  */
 #pragma once
 
-#include <optional>
 #include <raft/cluster/detail/kmeans.cuh>
 #include <raft/cluster/detail/kmeans_auto_find_k.cuh>
 #include <raft/cluster/kmeans_types.hpp>
@@ -24,6 +23,8 @@
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 
+#include <optional>
+
 namespace raft::cluster::kmeans {
 
 /**
diff --git a/cpp/include/raft/cluster/kmeans_balanced.cuh b/cpp/include/raft/cluster/kmeans_balanced.cuh
index 5c59f1393c..8cd7730814 100644
--- a/cpp/include/raft/cluster/kmeans_balanced.cuh
+++ b/cpp/include/raft/cluster/kmeans_balanced.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <raft/core/resource/device_memory_resource.hpp>
-#include <utility>
-
 #include <raft/cluster/detail/kmeans_balanced.cuh>
 #include <raft/core/mdarray.hpp>
+#include <raft/core/resource/device_memory_resource.hpp>
 #include <raft/util/cuda_utils.cuh>
 
+#include <utility>
+
 namespace raft::cluster::kmeans_balanced {
 
 /**
diff --git a/cpp/include/raft/comms/comms_test.hpp b/cpp/include/raft/comms/comms_test.hpp
index 3ceb2942a8..c2c07ebc8e 100644
--- a/cpp/include/raft/comms/comms_test.hpp
+++ b/cpp/include/raft/comms/comms_test.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,7 +18,6 @@
 
 #include <raft/comms/comms.hpp>
 #include <raft/comms/detail/test.hpp>
-
 #include <raft/core/resources.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/comms/detail/mpi_comms.hpp b/cpp/include/raft/comms/detail/mpi_comms.hpp
index 3342fec973..a07146052e 100644
--- a/cpp/include/raft/comms/detail/mpi_comms.hpp
+++ b/cpp/include/raft/comms/detail/mpi_comms.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,24 +16,24 @@
 
 #pragma once
 
-#include <cstdio>
-#include <memory>
-
-#include <unordered_map>
-#include <unordered_set>
-#include <utility>
-
-#include <mpi.h>
-#include <nccl.h>
-
 #include <raft/comms/comms.hpp>
 #include <raft/comms/detail/util.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_scalar.hpp>
 
+#include <mpi.h>
+#include <nccl.h>
+
+#include <cstdio>
+#include <memory>
+#include <unordered_map>
+#include <unordered_set>
+#include <utility>
+
 #define RAFT_MPI_TRY(call)                                                                    \
   do {                                                                                        \
     int status = call;                                                                        \
diff --git a/cpp/include/raft/comms/detail/std_comms.hpp b/cpp/include/raft/comms/detail/std_comms.hpp
index 323e408cab..6e7ff7106f 100644
--- a/cpp/include/raft/comms/detail/std_comms.hpp
+++ b/cpp/include/raft/comms/detail/std_comms.hpp
@@ -19,36 +19,31 @@
 #include <raft/comms/comms.hpp>
 #include <raft/comms/detail/ucp_helper.hpp>
 #include <raft/comms/detail/util.hpp>
-
-#include <raft/core/resources.hpp>
-#include <rmm/device_scalar.hpp>
-#include <rmm/device_uvector.hpp>
-
 #include <raft/core/error.hpp>
-
+#include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 
-#include <thrust/iterator/zip_iterator.h>
+#include <rmm/device_scalar.hpp>
+#include <rmm/device_uvector.hpp>
 
 #include <cuda_runtime.h>
+#include <thrust/iterator/zip_iterator.h>
 
+#include <nccl.h>
+#include <stdlib.h>
+#include <time.h>
 #include <ucp/api/ucp.h>
 #include <ucp/api/ucp_def.h>
 
-#include <nccl.h>
-
-#include <unordered_map>
-#include <unordered_set>
-#include <utility>
-
 #include <algorithm>
 #include <chrono>
 #include <cstdio>
 #include <exception>
 #include <memory>
-#include <stdlib.h>
 #include <thread>
-#include <time.h>
+#include <unordered_map>
+#include <unordered_set>
+#include <utility>
 
 namespace raft {
 namespace comms {
diff --git a/cpp/include/raft/comms/detail/test.hpp b/cpp/include/raft/comms/detail/test.hpp
index 876a17de1a..601c1858a3 100644
--- a/cpp/include/raft/comms/detail/test.hpp
+++ b/cpp/include/raft/comms/detail/test.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,15 +20,13 @@
 #include <raft/core/resource/comms.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
 #include <iostream>
 #include <numeric>
 
-#include <iostream>
-#include <numeric>
-
 namespace raft {
 namespace comms {
 namespace detail {
diff --git a/cpp/include/raft/comms/detail/ucp_helper.hpp b/cpp/include/raft/comms/detail/ucp_helper.hpp
index 9479bc24f9..5896248c1d 100644
--- a/cpp/include/raft/comms/detail/ucp_helper.hpp
+++ b/cpp/include/raft/comms/detail/ucp_helper.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include <raft/util/cudart_utils.hpp>
+
 #include <stdio.h>
 #include <ucp/api/ucp.h>
 #include <ucp/api/ucp_def.h>
diff --git a/cpp/include/raft/comms/detail/util.hpp b/cpp/include/raft/comms/detail/util.hpp
index 85084f0eae..c8cc9bf0bf 100644
--- a/cpp/include/raft/comms/detail/util.hpp
+++ b/cpp/include/raft/comms/detail/util.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,11 @@
 
 #pragma once
 
+#include <raft/core/error.hpp>
 #include <raft/core/interruptible.hpp>
 
 #include <nccl.h>
-#include <raft/core/error.hpp>
+
 #include <string>
 
 /**
diff --git a/cpp/include/raft/comms/std_comms.hpp b/cpp/include/raft/comms/std_comms.hpp
index df1687f93d..c81b19c9ba 100644
--- a/cpp/include/raft/comms/std_comms.hpp
+++ b/cpp/include/raft/comms/std_comms.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,17 +16,17 @@
 
 #pragma once
 
+#include <raft/comms/comms.hpp>
+#include <raft/comms/detail/std_comms.hpp>
 #include <raft/core/resource/comms.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
-#include <raft/comms/comms.hpp>
-#include <raft/comms/detail/std_comms.hpp>
-
-#include <iostream>
 #include <nccl.h>
 #include <ucp/api/ucp.h>
 
+#include <iostream>
+
 namespace raft {
 namespace comms {
 
diff --git a/cpp/include/raft/core/bitset.cuh b/cpp/include/raft/core/bitset.cuh
index 552c2e9ac5..53fd586ed2 100644
--- a/cpp/include/raft/core/bitset.cuh
+++ b/cpp/include/raft/core/bitset.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,6 +24,7 @@
 #include <raft/linalg/map.cuh>
 #include <raft/linalg/reduce.cuh>
 #include <raft/util/device_atomics.cuh>
+
 #include <thrust/for_each.h>
 
 namespace raft::core {
diff --git a/cpp/include/raft/core/comms.hpp b/cpp/include/raft/core/comms.hpp
index 463c17f2f6..7dc4800ed6 100644
--- a/cpp/include/raft/core/comms.hpp
+++ b/cpp/include/raft/core/comms.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,11 @@
 
 #pragma once
 
+#include <raft/core/error.hpp>
+
 #include <cuda_runtime.h>
+
 #include <memory>
-#include <raft/core/error.hpp>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/include/raft/core/cublas_macros.hpp b/cpp/include/raft/core/cublas_macros.hpp
index c782c8ded1..b69b121161 100644
--- a/cpp/include/raft/core/cublas_macros.hpp
+++ b/cpp/include/raft/core/cublas_macros.hpp
@@ -19,9 +19,10 @@
 
 #pragma once
 
-#include <cublas_v2.h>
 #include <raft/core/error.hpp>
 
+#include <cublas_v2.h>
+
 ///@todo: enable this once we have logger enabled
 // #include <cuml/common/logger.hpp>
 
diff --git a/cpp/include/raft/core/cusolver_macros.hpp b/cpp/include/raft/core/cusolver_macros.hpp
index 4477d32118..74a8b7c36c 100644
--- a/cpp/include/raft/core/cusolver_macros.hpp
+++ b/cpp/include/raft/core/cusolver_macros.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,6 +24,7 @@
 ///@todo: enable this once logging is enabled
 // #include <cuml/common/logger.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <type_traits>
 
 #define _CUSOLVER_ERR_TO_STR(err) \
diff --git a/cpp/include/raft/core/cusparse_macros.hpp b/cpp/include/raft/core/cusparse_macros.hpp
index 21a25ae28c..9058f4847d 100644
--- a/cpp/include/raft/core/cusparse_macros.hpp
+++ b/cpp/include/raft/core/cusparse_macros.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,8 +16,9 @@
 
 #pragma once
 
-#include <cusparse.h>
 #include <raft/core/error.hpp>
+
+#include <cusparse.h>
 ///@todo: enable this once logging is enabled
 // #include <cuml/common/logger.hpp>
 
diff --git a/cpp/include/raft/core/detail/copy.hpp b/cpp/include/raft/core/detail/copy.hpp
index 1a9a4d004d..04e74c4e58 100644
--- a/cpp/include/raft/core/detail/copy.hpp
+++ b/cpp/include/raft/core/detail/copy.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,6 @@
  */
 
 #pragma once
-#include <cstdio>
 #include <raft/core/cuda_support.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/error.hpp>
@@ -24,6 +23,8 @@
 #include <raft/core/mdspan.hpp>
 #include <raft/core/resource/stream_view.hpp>
 #include <raft/core/resources.hpp>
+
+#include <cstdio>
 #include <type_traits>
 #ifndef RAFT_DISABLE_CUDA
 #include <raft/core/cudart_utils.hpp>
diff --git a/cpp/include/raft/core/detail/fail_container_policy.hpp b/cpp/include/raft/core/detail/fail_container_policy.hpp
index ff36659f04..cf9d0887dd 100644
--- a/cpp/include/raft/core/detail/fail_container_policy.hpp
+++ b/cpp/include/raft/core/detail/fail_container_policy.hpp
@@ -19,6 +19,7 @@
 #include <raft/core/logger-macros.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/thirdparty/mdspan/include/experimental/mdspan>
+
 #include <stddef.h>
 
 namespace raft {
diff --git a/cpp/include/raft/core/detail/mdspan_util.cuh b/cpp/include/raft/core/detail/mdspan_util.cuh
index 6b2c90abcc..ded95c2f31 100644
--- a/cpp/include/raft/core/detail/mdspan_util.cuh
+++ b/cpp/include/raft/core/detail/mdspan_util.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,9 @@
  */
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
 #include <raft/thirdparty/mdspan/include/experimental/mdspan>
 
-#include <raft/core/detail/macros.hpp>
 #include <tuple>
 #include <utility>
 
diff --git a/cpp/include/raft/core/detail/nvtx.hpp b/cpp/include/raft/core/detail/nvtx.hpp
index f077a49b77..8afd1f16c6 100644
--- a/cpp/include/raft/core/detail/nvtx.hpp
+++ b/cpp/include/raft/core/detail/nvtx.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,10 +20,11 @@
 
 #ifdef NVTX_ENABLED
 
+#include <nvtx3/nvToolsExt.h>
+
 #include <cstdint>
 #include <cstdlib>
 #include <mutex>
-#include <nvtx3/nvToolsExt.h>
 #include <string>
 #include <type_traits>
 #include <unordered_map>
diff --git a/cpp/include/raft/core/detail/span.hpp b/cpp/include/raft/core/detail/span.hpp
index e6ccb8535c..5719fdfd2b 100644
--- a/cpp/include/raft/core/detail/span.hpp
+++ b/cpp/include/raft/core/detail/span.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,10 @@
  */
 #pragma once
 
-#include <limits>  // numeric_limits
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/mdspan_types.hpp>
+
+#include <limits>  // numeric_limits
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/core/device_container_policy.hpp b/cpp/include/raft/core/device_container_policy.hpp
index e8717d4c5e..8c6eff582b 100644
--- a/cpp/include/raft/core/device_container_policy.hpp
+++ b/cpp/include/raft/core/device_container_policy.hpp
@@ -22,14 +22,13 @@
  */
 #pragma once
 #ifndef RAFT_DISABLE_CUDA
-#include <raft/core/device_mdspan.hpp>
-#include <raft/util/cudart_utils.hpp>
-
 #include <raft/core/detail/span.hpp>  // dynamic_extent
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_device_accessor.hpp>
-
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/device_memory_resource.hpp>
diff --git a/cpp/include/raft/core/device_csr_matrix.hpp b/cpp/include/raft/core/device_csr_matrix.hpp
index da4ac117b1..1d23c8912d 100644
--- a/cpp/include/raft/core/device_csr_matrix.hpp
+++ b/cpp/include/raft/core/device_csr_matrix.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,6 +21,7 @@
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/core/sparse_types.hpp>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/core/device_mdarray.hpp b/cpp/include/raft/core/device_mdarray.hpp
index fe543c97dd..855642cd76 100644
--- a/cpp/include/raft/core/device_mdarray.hpp
+++ b/cpp/include/raft/core/device_mdarray.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,13 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/device_container_policy.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/device_mdspan.hpp b/cpp/include/raft/core/device_mdspan.hpp
index 7988bd3f6f..b4e9f8d1d7 100644
--- a/cpp/include/raft/core/device_mdspan.hpp
+++ b/cpp/include/raft/core/device_mdspan.hpp
@@ -16,11 +16,12 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/memory_type.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 template <typename AccessorPolicy>
diff --git a/cpp/include/raft/core/device_resources.hpp b/cpp/include/raft/core/device_resources.hpp
index cf06920a8c..366e387fdd 100644
--- a/cpp/include/raft/core/device_resources.hpp
+++ b/cpp/include/raft/core/device_resources.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,24 +19,7 @@
 
 #pragma once
 
-#include <memory>
-#include <mutex>
-#include <optional>
-#include <string>
-#include <unordered_map>
-#include <utility>
-#include <vector>
-
-#include <cublas_v2.h>
-#include <cuda_runtime.h>
-#include <cusolverDn.h>
-#include <cusolverSp.h>
-#include <cusparse.h>
-
 #include <raft/core/comms.hpp>
-#include <rmm/cuda_stream_pool.hpp>
-#include <rmm/exec_policy.hpp>
-
 #include <raft/core/resource/comms.hpp>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_event.hpp>
@@ -52,6 +35,24 @@
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 
+#include <rmm/cuda_stream_pool.hpp>
+#include <rmm/exec_policy.hpp>
+
+#include <cuda_runtime.h>
+
+#include <cublas_v2.h>
+#include <cusolverDn.h>
+#include <cusolverSp.h>
+#include <cusparse.h>
+
+#include <memory>
+#include <mutex>
+#include <optional>
+#include <string>
+#include <unordered_map>
+#include <utility>
+#include <vector>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/device_resources_manager.hpp b/cpp/include/raft/core/device_resources_manager.hpp
index 75c9428ff7..1c4bee15a6 100644
--- a/cpp/include/raft/core/device_resources_manager.hpp
+++ b/cpp/include/raft/core/device_resources_manager.hpp
@@ -15,17 +15,19 @@
  */
 
 #pragma once
-#include <algorithm>
-#include <memory>
-#include <optional>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/device_setter.hpp>
+
 #include <rmm/cuda_device.hpp>
 #include <rmm/cuda_stream.hpp>
 #include <rmm/cuda_stream_pool.hpp>
 #include <rmm/mr/device/cuda_memory_resource.hpp>
 #include <rmm/mr/device/per_device_resource.hpp>
 
+#include <algorithm>
+#include <memory>
+#include <optional>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/device_setter.hpp b/cpp/include/raft/core/device_setter.hpp
index 42049102aa..76ac65de8e 100644
--- a/cpp/include/raft/core/device_setter.hpp
+++ b/cpp/include/raft/core/device_setter.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,10 @@
  */
 
 #pragma once
-#include <cuda_runtime_api.h>
 #include <raft/core/logger.hpp>
 #include <raft/util/cuda_rt_essentials.hpp>
+
+#include <cuda_runtime_api.h>
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/error.hpp b/cpp/include/raft/core/error.hpp
index 6533e9c3c6..63b7d5e626 100644
--- a/cpp/include/raft/core/error.hpp
+++ b/cpp/include/raft/core/error.hpp
@@ -33,6 +33,7 @@
 #ifdef ENABLE_COLLECT_CALLSTACK
 #include <cxxabi.h>
 #include <execinfo.h>
+
 #include <sstream>
 #endif
 
diff --git a/cpp/include/raft/core/host_container_policy.hpp b/cpp/include/raft/core/host_container_policy.hpp
index 0192436934..fa44f1b138 100644
--- a/cpp/include/raft/core/host_container_policy.hpp
+++ b/cpp/include/raft/core/host_container_policy.hpp
@@ -23,6 +23,7 @@
 #pragma once
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/resources.hpp>
+
 #include <vector>
 
 namespace raft {
diff --git a/cpp/include/raft/core/host_csr_matrix.hpp b/cpp/include/raft/core/host_csr_matrix.hpp
index f32ff1dc00..e3cea3cd27 100644
--- a/cpp/include/raft/core/host_csr_matrix.hpp
+++ b/cpp/include/raft/core/host_csr_matrix.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,6 +20,7 @@
 #include <raft/core/host_span.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/core/sparse_types.hpp>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/core/host_device_accessor.hpp b/cpp/include/raft/core/host_device_accessor.hpp
index 7cb2aaf487..cbcfa976b2 100644
--- a/cpp/include/raft/core/host_device_accessor.hpp
+++ b/cpp/include/raft/core/host_device_accessor.hpp
@@ -16,6 +16,7 @@
 
 #pragma once
 #include <raft/core/memory_type.hpp>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/core/host_mdarray.hpp b/cpp/include/raft/core/host_mdarray.hpp
index 02c8914ca1..3020cde32d 100644
--- a/cpp/include/raft/core/host_mdarray.hpp
+++ b/cpp/include/raft/core/host_mdarray.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,12 @@
 
 #pragma once
 
-#include <cstdint>
+#include <raft/core/host_container_policy.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/mdarray.hpp>
 #include <raft/core/resources.hpp>
 
-#include <raft/core/host_container_policy.hpp>
-#include <raft/core/mdarray.hpp>
+#include <cstdint>
 
 namespace raft {
 /**
diff --git a/cpp/include/raft/core/host_mdspan.hpp b/cpp/include/raft/core/host_mdspan.hpp
index d5f431f4a2..ef0a9a2c0f 100644
--- a/cpp/include/raft/core/host_mdspan.hpp
+++ b/cpp/include/raft/core/host_mdspan.hpp
@@ -16,11 +16,11 @@
 
 #pragma once
 
-#include <cstdint>
+#include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/memory_type.hpp>
 
-#include <raft/core/host_device_accessor.hpp>
+#include <cstdint>
 
 namespace raft {
 
diff --git a/cpp/include/raft/core/interruptible.hpp b/cpp/include/raft/core/interruptible.hpp
index 10ab22f820..4ec1067e00 100644
--- a/cpp/include/raft/core/interruptible.hpp
+++ b/cpp/include/raft/core/interruptible.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,12 +19,14 @@
 
 #pragma once
 
-#include <memory>
-#include <mutex>
-#include <optional>
 #include <raft/core/error.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
+
+#include <memory>
+#include <mutex>
+#include <optional>
 #include <thread>
 #include <unordered_map>
 
diff --git a/cpp/include/raft/core/kvp.hpp b/cpp/include/raft/core/kvp.hpp
index 610945a76c..13686c0c9b 100644
--- a/cpp/include/raft/core/kvp.hpp
+++ b/cpp/include/raft/core/kvp.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,8 +19,9 @@
 #include <raft/core/detail/macros.hpp>
 
 #ifdef _RAFT_HAS_CUDA
-#include <cub/cub.cuh>
 #include <raft/util/cuda_utils.cuh>  // raft::shfl_xor
+
+#include <cub/cub.cuh>
 #endif
 namespace raft {
 /**
diff --git a/cpp/include/raft/core/logger-ext.hpp b/cpp/include/raft/core/logger-ext.hpp
index 04a6a4d060..73fe463aba 100644
--- a/cpp/include/raft/core/logger-ext.hpp
+++ b/cpp/include/raft/core/logger-ext.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,10 +15,11 @@
  */
 #pragma once
 
-#include <memory>                       // std::unique_ptr
 #include <raft/core/detail/macros.hpp>  // RAFT_INLINE_CONDITIONAL
-#include <string>                       // std::string
-#include <unordered_map>                // std::unordered_map
+
+#include <memory>         // std::unique_ptr
+#include <string>         // std::string
+#include <unordered_map>  // std::unordered_map
 
 namespace raft {
 
diff --git a/cpp/include/raft/core/logger-inl.hpp b/cpp/include/raft/core/logger-inl.hpp
index fcfa1f1333..ea5f4ea26e 100644
--- a/cpp/include/raft/core/logger-inl.hpp
+++ b/cpp/include/raft/core/logger-inl.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,19 +15,16 @@
  */
 #pragma once
 
+#include "logger-macros.hpp"
+
 #include <stdarg.h>
 
 #include <algorithm>
-
 #include <memory>
 #include <mutex>
 #include <sstream>
 #include <string>
 #include <unordered_map>
-
-#include <stdarg.h>
-
-#include "logger-macros.hpp"
 // The logger-ext.hpp file contains the class declaration of the logger class.
 // In this case, it is okay to include the logger-ext.hpp file because it
 // contains no RAFT_EXPLICIT template instantiations.
@@ -35,7 +32,8 @@
 
 #define SPDLOG_HEADER_ONLY
 #include <raft/core/detail/callback_sink.hpp>
-#include <raft/core/detail/macros.hpp>        // RAFT_INLINE_CONDITIONAL
+#include <raft/core/detail/macros.hpp>  // RAFT_INLINE_CONDITIONAL
+
 #include <spdlog/sinks/stdout_color_sinks.h>  // NOLINT
 #include <spdlog/spdlog.h>                    // NOLINT
 
diff --git a/cpp/include/raft/core/logger.hpp b/cpp/include/raft/core/logger.hpp
index 59968ff5e5..e64a0db257 100644
--- a/cpp/include/raft/core/logger.hpp
+++ b/cpp/include/raft/core/logger.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,8 @@
  */
 #pragma once
 
-#include "logger-macros.hpp"
-
 #include "logger-ext.hpp"
+#include "logger-macros.hpp"
 
 #if !defined(RAFT_COMPILED)
 #include "logger-inl.hpp"
diff --git a/cpp/include/raft/core/managed_container_policy.hpp b/cpp/include/raft/core/managed_container_policy.hpp
index f4e26c6ef1..889e9383e6 100644
--- a/cpp/include/raft/core/managed_container_policy.hpp
+++ b/cpp/include/raft/core/managed_container_policy.hpp
@@ -15,15 +15,14 @@
  */
 #pragma once
 #ifndef RAFT_DISABLE_CUDA
+#include <raft/core/detail/span.hpp>  // dynamic_extent
 #include <raft/core/device_container_policy.hpp>
 #include <raft/core/device_mdspan.hpp>
-#include <raft/util/cudart_utils.hpp>
-
-#include <raft/core/detail/span.hpp>  // dynamic_extent
 #include <raft/core/host_device_accessor.hpp>
-
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/util/cudart_utils.hpp>
+
 #include <rmm/mr/device/managed_memory_resource.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/core/managed_mdarray.hpp b/cpp/include/raft/core/managed_mdarray.hpp
index c1438d941d..021fa11321 100644
--- a/cpp/include/raft/core/managed_mdarray.hpp
+++ b/cpp/include/raft/core/managed_mdarray.hpp
@@ -16,12 +16,13 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/managed_container_policy.hpp>
 #include <raft/core/managed_mdspan.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/managed_mdspan.hpp b/cpp/include/raft/core/managed_mdspan.hpp
index 9c2976ec6b..b421982df4 100644
--- a/cpp/include/raft/core/managed_mdspan.hpp
+++ b/cpp/include/raft/core/managed_mdspan.hpp
@@ -16,11 +16,12 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/memory_type.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 template <typename AccessorPolicy>
diff --git a/cpp/include/raft/core/math.hpp b/cpp/include/raft/core/math.hpp
index 809b2948e7..9ce768bf40 100644
--- a/cpp/include/raft/core/math.hpp
+++ b/cpp/include/raft/core/math.hpp
@@ -16,12 +16,12 @@
 
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
+
 #include <algorithm>
 #include <cmath>
 #include <type_traits>
 
-#include <raft/core/detail/macros.hpp>
-
 #if defined(_RAFT_HAS_CUDA)
 #include <cuda_bf16.h>
 #include <cuda_fp16.h>
diff --git a/cpp/include/raft/core/mdarray.hpp b/cpp/include/raft/core/mdarray.hpp
index 2cdeb36fc8..0fa8e98df9 100644
--- a/cpp/include/raft/core/mdarray.hpp
+++ b/cpp/include/raft/core/mdarray.hpp
@@ -6,7 +6,7 @@
  */
 
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -22,8 +22,6 @@
  */
 #pragma once
 
-#include <stddef.h>
-
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
@@ -31,6 +29,8 @@
 #include <raft/core/memory_type.hpp>
 #include <raft/core/resources.hpp>
 
+#include <stddef.h>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/mdbuffer.cuh b/cpp/include/raft/core/mdbuffer.cuh
index 18533ce882..d9557f8c55 100644
--- a/cpp/include/raft/core/mdbuffer.cuh
+++ b/cpp/include/raft/core/mdbuffer.cuh
@@ -15,9 +15,6 @@
  */
 
 #pragma once
-#include <algorithm>
-#include <execution>
-#include <optional>
 #include <raft/core/cuda_support.hpp>
 #include <raft/core/detail/copy.hpp>
 #include <raft/core/device_container_policy.hpp>
@@ -31,12 +28,17 @@
 #include <raft/core/pinned_container_policy.hpp>
 #include <raft/core/stream_view.hpp>
 #include <raft/util/variant_utils.hpp>
+
+#include <algorithm>
+#include <execution>
+#include <optional>
 #include <type_traits>
 #include <utility>
 #include <variant>
 #ifndef RAFT_DISABLE_CUDA
 #include <raft/core/copy.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <thrust/device_ptr.h>
 #else
 #include <raft/core/copy.hpp>
diff --git a/cpp/include/raft/core/mdspan.hpp b/cpp/include/raft/core/mdspan.hpp
index 15414b9af3..45bebaec2f 100644
--- a/cpp/include/raft/core/mdspan.hpp
+++ b/cpp/include/raft/core/mdspan.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,12 @@
  */
 #pragma once
 
-#include <raft/core/error.hpp>
-#include <raft/core/mdspan_types.hpp>
-#include <raft/core/memory_type.hpp>
-
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/detail/mdspan_util.cuh>
+#include <raft/core/error.hpp>
 #include <raft/core/host_device_accessor.hpp>
-
+#include <raft/core/mdspan_types.hpp>
+#include <raft/core/memory_type.hpp>
 #include <raft/thirdparty/mdspan/include/experimental/mdspan>
 
 namespace raft {
diff --git a/cpp/include/raft/core/memory_type.hpp b/cpp/include/raft/core/memory_type.hpp
index 7849cd67ab..0334cb5059 100644
--- a/cpp/include/raft/core/memory_type.hpp
+++ b/cpp/include/raft/core/memory_type.hpp
@@ -17,8 +17,10 @@
 #include <cstdint>
 #include <optional>
 #ifndef RAFT_DISABLE_CUDA
-#include <cuda_runtime.h>
 #include <raft/util/cuda_rt_essentials.hpp>
+
+#include <cuda_runtime.h>
+
 #include <type_traits>
 #else
 #include <raft/core/logger.hpp>
diff --git a/cpp/include/raft/core/nvtx.hpp b/cpp/include/raft/core/nvtx.hpp
index 57338c32c7..26e3a383f2 100644
--- a/cpp/include/raft/core/nvtx.hpp
+++ b/cpp/include/raft/core/nvtx.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,10 @@
 
 #pragma once
 
-#include <optional>
 #include <raft/core/detail/nvtx.hpp>
 
+#include <optional>
+
 /**
  * \section Usage
  *
diff --git a/cpp/include/raft/core/operators.hpp b/cpp/include/raft/core/operators.hpp
index 7acc907c49..e42801fe32 100644
--- a/cpp/include/raft/core/operators.hpp
+++ b/cpp/include/raft/core/operators.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,15 @@
 
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
+#include <raft/core/math.hpp>
+
 #include <algorithm>
 #include <cmath>
 #include <tuple>
 #include <type_traits>
 #include <utility>
 
-#include <raft/core/detail/macros.hpp>
-#include <raft/core/math.hpp>
-
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/pinned_container_policy.hpp b/cpp/include/raft/core/pinned_container_policy.hpp
index 51451deadb..4870e2c5dc 100644
--- a/cpp/include/raft/core/pinned_container_policy.hpp
+++ b/cpp/include/raft/core/pinned_container_policy.hpp
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 #pragma once
-#include <cstddef>
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/resources.hpp>
+
+#include <cstddef>
 #ifndef RAFT_DISABLE_CUDA
 #include <thrust/host_vector.h>
 #include <thrust/mr/allocator.h>
diff --git a/cpp/include/raft/core/pinned_mdarray.hpp b/cpp/include/raft/core/pinned_mdarray.hpp
index 72b8d52e0d..f289062a02 100644
--- a/cpp/include/raft/core/pinned_mdarray.hpp
+++ b/cpp/include/raft/core/pinned_mdarray.hpp
@@ -16,12 +16,13 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/pinned_container_policy.hpp>
 #include <raft/core/pinned_mdspan.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/core/pinned_mdspan.hpp b/cpp/include/raft/core/pinned_mdspan.hpp
index e764101d1c..d03433a8db 100644
--- a/cpp/include/raft/core/pinned_mdspan.hpp
+++ b/cpp/include/raft/core/pinned_mdspan.hpp
@@ -16,11 +16,12 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/memory_type.hpp>
 
+#include <cstdint>
+
 namespace raft {
 
 template <typename AccessorPolicy>
diff --git a/cpp/include/raft/core/resource/cublas_handle.hpp b/cpp/include/raft/core/resource/cublas_handle.hpp
index 306a641f34..1c86f3e93f 100644
--- a/cpp/include/raft/core/resource/cublas_handle.hpp
+++ b/cpp/include/raft/core/resource/cublas_handle.hpp
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cublas_v2.h>
 #include <raft/core/cublas_macros.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cublas_v2.h>
+
 namespace raft::resource {
 
 class cublas_resource : public resource {
diff --git a/cpp/include/raft/core/resource/cublaslt_handle.hpp b/cpp/include/raft/core/resource/cublaslt_handle.hpp
index a61e56bd91..4653a55e16 100644
--- a/cpp/include/raft/core/resource/cublaslt_handle.hpp
+++ b/cpp/include/raft/core/resource/cublaslt_handle.hpp
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <cublasLt.h>
 #include <raft/core/cublas_macros.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cublasLt.h>
+
 #include <memory>
 
 namespace raft::resource {
diff --git a/cpp/include/raft/core/resource/cuda_event.hpp b/cpp/include/raft/core/resource/cuda_event.hpp
index 4859d95ee9..db8e40925c 100644
--- a/cpp/include/raft/core/resource/cuda_event.hpp
+++ b/cpp/include/raft/core/resource/cuda_event.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource {
 
 class cuda_event_resource : public resource {
diff --git a/cpp/include/raft/core/resource/cuda_stream.hpp b/cpp/include/raft/core/resource/cuda_stream.hpp
index 6120a4b75c..68b5c055fe 100644
--- a/cpp/include/raft/core/resource/cuda_stream.hpp
+++ b/cpp/include/raft/core/resource/cuda_stream.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource {
 class cuda_stream_resource : public resource {
  public:
diff --git a/cpp/include/raft/core/resource/cuda_stream_pool.hpp b/cpp/include/raft/core/resource/cuda_stream_pool.hpp
index 7ed356485c..c15a640909 100644
--- a/cpp/include/raft/core/resource/cuda_stream_pool.hpp
+++ b/cpp/include/raft/core/resource/cuda_stream_pool.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,15 +15,16 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/resource/cuda_event.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/detail/stream_sync_event.hpp>
-
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
+
 #include <rmm/cuda_stream_pool.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource {
 
 class cuda_stream_pool_resource : public resource {
diff --git a/cpp/include/raft/core/resource/cusolver_dn_handle.hpp b/cpp/include/raft/core/resource/cusolver_dn_handle.hpp
index 7a33e2dd2a..a0698d9b82 100644
--- a/cpp/include/raft/core/resource/cusolver_dn_handle.hpp
+++ b/cpp/include/raft/core/resource/cusolver_dn_handle.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,15 @@
 #pragma once
 
 #include "cuda_stream.hpp"
-#include <cusolverDn.h>
+
 #include <raft/core/cusolver_macros.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 
+#include <cusolverDn.h>
+
 namespace raft::resource {
 
 /**
diff --git a/cpp/include/raft/core/resource/cusolver_sp_handle.hpp b/cpp/include/raft/core/resource/cusolver_sp_handle.hpp
index 61fd95b44f..f28724632c 100644
--- a/cpp/include/raft/core/resource/cusolver_sp_handle.hpp
+++ b/cpp/include/raft/core/resource/cusolver_sp_handle.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cusolverSp.h>
 #include <raft/core/cusolver_macros.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cusolverSp.h>
+
 namespace raft::resource {
 
 /**
diff --git a/cpp/include/raft/core/resource/cusparse_handle.hpp b/cpp/include/raft/core/resource/cusparse_handle.hpp
index 9893ed2f86..f019b41a52 100644
--- a/cpp/include/raft/core/resource/cusparse_handle.hpp
+++ b/cpp/include/raft/core/resource/cusparse_handle.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/core/cusparse_macros.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cusparse_v2.h>
+
 namespace raft::resource {
 class cusparse_resource : public resource {
  public:
diff --git a/cpp/include/raft/core/resource/detail/stream_sync_event.hpp b/cpp/include/raft/core/resource/detail/stream_sync_event.hpp
index 1d02fef20d..8429c7667d 100644
--- a/cpp/include/raft/core/resource/detail/stream_sync_event.hpp
+++ b/cpp/include/raft/core/resource/detail/stream_sync_event.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/resource/cuda_event.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource::detail {
 
 /**
diff --git a/cpp/include/raft/core/resource/device_id.hpp b/cpp/include/raft/core/resource/device_id.hpp
index b55e56ca45..570d815780 100644
--- a/cpp/include/raft/core/resource/device_id.hpp
+++ b/cpp/include/raft/core/resource/device_id.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource {
 
 class device_id_resource : public resource {
diff --git a/cpp/include/raft/core/resource/device_properties.hpp b/cpp/include/raft/core/resource/device_properties.hpp
index c3b0b8f2b9..a87c29f709 100644
--- a/cpp/include/raft/core/resource/device_properties.hpp
+++ b/cpp/include/raft/core/resource/device_properties.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cuda_runtime.h>
 #include <raft/core/resource/device_id.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
+
 namespace raft::resource {
 
 class device_properties_resource : public resource {
diff --git a/cpp/include/raft/core/resource/thrust_policy.hpp b/cpp/include/raft/core/resource/thrust_policy.hpp
index 78c04ce875..f81898be8a 100644
--- a/cpp/include/raft/core/resource/thrust_policy.hpp
+++ b/cpp/include/raft/core/resource/thrust_policy.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,6 +18,7 @@
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/resource_types.hpp>
 #include <raft/core/resources.hpp>
+
 #include <rmm/exec_policy.hpp>
 namespace raft::resource {
 class thrust_policy_resource : public resource {
diff --git a/cpp/include/raft/core/resources.hpp b/cpp/include/raft/core/resources.hpp
index d5bd176d50..b0827d8e11 100644
--- a/cpp/include/raft/core/resources.hpp
+++ b/cpp/include/raft/core/resources.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,12 @@
 #pragma once
 
 #include "resource/resource_types.hpp"
-#include <algorithm>
-#include <mutex>
+
 #include <raft/core/error.hpp>  // RAFT_EXPECTS
 #include <raft/core/logger.hpp>
+
+#include <algorithm>
+#include <mutex>
 #include <string>
 #include <vector>
 
diff --git a/cpp/include/raft/core/span.hpp b/cpp/include/raft/core/span.hpp
index d77e0fcb40..965600f490 100644
--- a/cpp/include/raft/core/span.hpp
+++ b/cpp/include/raft/core/span.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,13 @@
  */
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
+#include <raft/core/detail/span.hpp>
+#include <raft/core/mdspan_types.hpp>
+
 #include <cassert>
 #include <cinttypes>  // size_t
 #include <cstddef>    // std::byte
-#include <raft/core/mdspan_types.hpp>
-
-#include <raft/core/detail/macros.hpp>
-#include <raft/core/detail/span.hpp>
 
 // TODO (cjnolet): Remove thrust dependencies here so host_span can be used without CUDA Toolkit
 // being installed. Reference: https://github.com/rapidsai/raft/issues/812.
@@ -29,6 +29,7 @@
 #include <thrust/functional.h>
 #include <thrust/host_vector.h>  // _RAFT_HOST_DEVICE
 #include <thrust/iterator/reverse_iterator.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/core/stream_view.hpp b/cpp/include/raft/core/stream_view.hpp
index 128050c414..3a42ee949e 100644
--- a/cpp/include/raft/core/stream_view.hpp
+++ b/cpp/include/raft/core/stream_view.hpp
@@ -19,6 +19,7 @@
 #include <raft/core/logger.hpp>
 #ifndef RAFT_DISABLE_CUDA
 #include <raft/core/interruptible.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 #endif
 
diff --git a/cpp/include/raft/core/temporary_device_buffer.hpp b/cpp/include/raft/core/temporary_device_buffer.hpp
index 358eeab861..8b07b51854 100644
--- a/cpp/include/raft/core/temporary_device_buffer.hpp
+++ b/cpp/include/raft/core/temporary_device_buffer.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,8 +18,8 @@
 
 #include "device_mdarray.hpp"
 #include "device_mdspan.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/cudart_utils.hpp>
 
 #include <variant>
diff --git a/cpp/include/raft/distance/detail/distance.cuh b/cpp/include/raft/distance/detail/distance.cuh
index b6885808ce..a5c8c0ef4b 100644
--- a/cpp/include/raft/distance/detail/distance.cuh
+++ b/cpp/include/raft/distance/detail/distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -27,6 +27,7 @@
 #include <raft/linalg/norm.cuh>
 #include <raft/linalg/reduce.cuh>
 #include <raft/linalg/unary_op.cuh>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h b/cpp/include/raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h
index ac20578083..186715851b 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h
@@ -67,30 +67,26 @@ reduction
 #include <cuda/std/utility>
 #else
 #include <assert.h>
+
 #include <utility>
 #endif
 
 #include <cutlass/aligned_buffer.h>
 #include <cutlass/array.h>
 #include <cutlass/cutlass.h>
+#include <cutlass/epilogue/threadblock/epilogue_base.h>
+#include <cutlass/epilogue/threadblock/predicated_tile_iterator.h>
 #include <cutlass/fast_math.h>
 #include <cutlass/functional.h>
+#include <cutlass/gemm/gemm.h>
 #include <cutlass/layout/tensor.h>
 #include <cutlass/layout/vector.h>
 #include <cutlass/numeric_conversion.h>
 #include <cutlass/numeric_types.h>
 #include <cutlass/tensor_coord.h>
-
-#include <cutlass/gemm/gemm.h>
-
 #include <cutlass/transform/pitch_linear_thread_map.h>
 #include <cutlass/transform/threadblock/regular_tile_iterator.h>
 
-#include <cutlass/epilogue/threadblock/epilogue_base.h>
-#include <cutlass/epilogue/threadblock/predicated_tile_iterator.h>
-
-#include <cutlass/numeric_types.h>
-
 /////////////////////////////////////////////////////////////////////////////////////////////////
 
 namespace cutlass {
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/cutlass_base.cuh b/cpp/include/raft/distance/detail/fused_distance_nn/cutlass_base.cuh
index be1c8891f5..3f359c2f54 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/cutlass_base.cuh
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/cutlass_base.cuh
@@ -43,6 +43,17 @@
 #include <raft/util/cudart_utils.hpp>   // getMultiProcessorCount
 #include <raft/util/cutlass_utils.cuh>  // RAFT_CUTLASS_TRY
 
+#include <rmm/device_uvector.hpp>
+
+#include <cutlass/cutlass.h>
+#include <cutlass/gemm/device/gemm.h>
+#include <cutlass/gemm/device/gemm_grouped.h>
+#include <cutlass/gemm/device/gemm_universal_adapter.h>
+#include <cutlass/layout/matrix.h>
+#include <cutlass/layout/tensor.h>
+#include <cutlass/matrix_coord.h>
+#include <cutlass/tensor_view.h>
+
 namespace raft {
 namespace distance {
 namespace detail {
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/epilogue.cuh b/cpp/include/raft/distance/detail/fused_distance_nn/epilogue.cuh
index 8a0bea3469..22848d30b0 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/epilogue.cuh
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/epilogue.cuh
@@ -29,7 +29,7 @@
  *
  **************************************************************************************************/
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -60,19 +60,17 @@ operation.
 
 #pragma once
 
+#include <raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h>
+#include <raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h>
+#include <raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h>
+
 #include <cutlass/array.h>
 #include <cutlass/cutlass.h>
-#include <cutlass/numeric_types.h>
-
-#include <cutlass/gemm/gemm.h>
-
 #include <cutlass/epilogue/threadblock/default_epilogue_tensor_op.h>
 #include <cutlass/epilogue/threadblock/default_epilogue_volta_tensor_op.h>
 #include <cutlass/epilogue/threadblock/epilogue.h>
-#include <raft/distance/detail/fused_distance_nn/custom_epilogue_with_broadcast.h>
-
-#include <raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h>
-#include <raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h>
+#include <cutlass/gemm/gemm.h>
+#include <cutlass/numeric_types.h>
 
 ////////////////////////////////////////////////////////////////////////////////
 
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/epilogue_elementwise.cuh b/cpp/include/raft/distance/detail/fused_distance_nn/epilogue_elementwise.cuh
index d65d2df4a4..993cf2f086 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/epilogue_elementwise.cuh
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/epilogue_elementwise.cuh
@@ -58,6 +58,7 @@
 
 #include <cutlass/array.h>
 #include <cutlass/cutlass.h>
+#include <cutlass/epilogue/thread/activation.h>
 #include <cutlass/functional.h>
 #include <cutlass/numeric_conversion.h>
 #include <cutlass/numeric_types.h>
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/gemm.h b/cpp/include/raft/distance/detail/fused_distance_nn/gemm.h
index 3da8b3ee3d..42de4860a0 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/gemm.h
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/gemm.h
@@ -29,7 +29,7 @@
  *
  **************************************************************************************************/
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -46,15 +46,14 @@
 
 #pragma once
 
-#include <cutlass/cutlass.h>
+#include <raft/distance/detail/fused_distance_nn/epilogue.cuh>
+#include <raft/distance/detail/fused_distance_nn/persistent_gemm.h>
 
+#include <cutlass/cutlass.h>
 #include <cutlass/gemm/kernel/default_gemm_universal.h>
 #include <cutlass/layout/matrix.h>
 #include <cutlass/layout/tensor.h>
 
-#include <raft/distance/detail/fused_distance_nn/epilogue.cuh>
-#include <raft/distance/detail/fused_distance_nn/persistent_gemm.h>
-
 /////////////////////////////////////////////////////////////////////////////////////////////////
 
 namespace cutlass {
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/persistent_gemm.h b/cpp/include/raft/distance/detail/fused_distance_nn/persistent_gemm.h
index 223af7eb58..f1a7c728e9 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/persistent_gemm.h
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/persistent_gemm.h
@@ -60,12 +60,11 @@ This file contains heavily customized version of GemmGrouped from CUTLASS 2.10.0
 #include <cutlass/cutlass.h>
 #include <cutlass/fast_math.h>
 #include <cutlass/gemm/gemm.h>
-#include <cutlass/matrix_coord.h>
-#include <cutlass/semaphore.h>
-
 #include <cutlass/gemm/kernel/gemm_grouped_problem_visitor.h>
 #include <cutlass/gemm/kernel/gemm_transpose_operands.h>
 #include <cutlass/layout/matrix.h>
+#include <cutlass/matrix_coord.h>
+#include <cutlass/semaphore.h>
 #include <cutlass/trace.h>
 
 /////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h b/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h
index 14c09f6aef..794cd5eb63 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_normvec_smem.h
@@ -29,7 +29,7 @@
  *
  **************************************************************************************************/
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -58,6 +58,8 @@ This file contains a customized version of PredicatedTileIterator from CUTLASS 2
 */
 #pragma once
 
+#include <raft/util/device_loads_stores.cuh>
+
 #include <cutlass/arch/arch.h>
 #include <cutlass/arch/memory.h>
 #include <cutlass/array.h>
@@ -71,8 +73,6 @@ This file contains a customized version of PredicatedTileIterator from CUTLASS 2
 #include <cutlass/tensor_ref.h>
 #include <cutlass/transform/pitch_linear_thread_map.h>
 
-#include <raft/util/device_loads_stores.cuh>
-
 ////////////////////////////////////////////////////////////////////////////////
 
 namespace cutlass {
diff --git a/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h b/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h
index 5ceb0dbaf5..d61018593f 100644
--- a/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h
+++ b/cpp/include/raft/distance/detail/fused_distance_nn/predicated_tile_iterator_reduced_vec.h
@@ -64,6 +64,7 @@ gmem.
 
 #include <cooperative_groups.h>
 #include <cooperative_groups/reduce.h>
+
 #include <cutlass/arch/arch.h>
 #include <cutlass/arch/memory.h>
 #include <cutlass/array.h>
diff --git a/cpp/include/raft/distance/detail/fused_l2_nn.cuh b/cpp/include/raft/distance/detail/fused_l2_nn.cuh
new file mode 100644
index 0000000000..5e62045305
--- /dev/null
+++ b/cpp/include/raft/distance/detail/fused_l2_nn.cuh
@@ -0,0 +1,386 @@
+/*
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#pragma once
+
+#include <raft/core/kvp.hpp>                             // raft::KeyValuePair
+#include <raft/core/operators.hpp>                       // raft::identity_op
+#include <raft/distance/detail/distance_ops/l2_exp.cuh>  // ops::l2_exp_distance_op
+#include <raft/distance/detail/fused_distance_nn/cutlass_base.cuh>
+#include <raft/distance/detail/pairwise_distance_base.cuh>  // PairwiseDistances
+#include <raft/linalg/contractions.cuh>                     // Policy
+#include <raft/util/arch.cuh>                               // raft::util::arch::SM_*
+#include <raft/util/cuda_utils.cuh>                         // raft::ceildiv, raft::shfl
+
+#include <cstddef>  // size_t
+#include <limits>   // std::numeric_limits
+
+namespace raft {
+namespace distance {
+
+namespace detail {
+
+template <typename LabelT, typename DataT>
+struct KVPMinReduceImpl {
+  typedef raft::KeyValuePair<LabelT, DataT> KVP;
+  DI KVP operator()(LabelT rit, const KVP& a, const KVP& b) { return b.value < a.value ? b : a; }
+  DI KVP operator()(const KVP& a, const KVP& b) { return b.value < a.value ? b : a; }
+
+};  // KVPMinReduce
+
+template <typename LabelT, typename DataT>
+struct MinAndDistanceReduceOpImpl {
+  typedef typename raft::KeyValuePair<LabelT, DataT> KVP;
+  DI void operator()(LabelT rid, KVP* out, const KVP& other) const
+  {
+    if (other.value < out->value) {
+      out->key   = other.key;
+      out->value = other.value;
+    }
+  }
+
+  DI void operator()(LabelT rid, DataT* out, const KVP& other) const
+  {
+    if (other.value < *out) { *out = other.value; }
+  }
+
+  DI void operator()(LabelT rid, DataT* out, const DataT& other) const
+  {
+    if (other < *out) { *out = other; }
+  }
+
+  DI void init(DataT* out, DataT maxVal) const { *out = maxVal; }
+  DI void init(KVP* out, DataT maxVal) const { out->value = maxVal; }
+
+  DI void init_key(DataT& out, LabelT idx) const { return; }
+  DI void init_key(KVP& out, LabelT idx) const { out.key = idx; }
+
+  DI DataT get_value(KVP& out) const
+  {
+    return out.value;
+    ;
+  }
+  DI DataT get_value(DataT& out) const { return out; }
+};
+
+template <typename LabelT, typename DataT>
+struct MinReduceOpImpl {
+  typedef typename raft::KeyValuePair<LabelT, DataT> KVP;
+  DI void operator()(LabelT rid, DataT* out, const KVP& other)
+  {
+    if (other.value < *out) { *out = other.value; }
+  }
+
+  DI void init(DataT* out, DataT maxVal) { *out = maxVal; }
+};
+
+template <typename DataT, typename OutT, typename IdxT, typename ReduceOpT>
+RAFT_KERNEL initKernel(OutT* min, IdxT m, DataT maxVal, ReduceOpT redOp)
+{
+  auto tid = IdxT(blockIdx.x) * blockDim.x + threadIdx.x;
+  if (tid < m) { redOp.init(min + tid, maxVal); }
+}
+
+template <typename DataT, typename OutT, typename IdxT, typename ReduceOpT>
+void initialize(OutT* min, IdxT m, DataT maxVal, ReduceOpT redOp, cudaStream_t stream)
+{
+  auto blks = raft::ceildiv(m, 256);
+  initKernel<DataT, OutT, IdxT><<<blks, 256, 0, stream>>>(min, m, maxVal, redOp);
+}
+
+// TODO: specialize this function for MinAndDistanceReduceOp<int, float>
+// with atomicCAS of 64 bit which will eliminate mutex and shfls
+template <typename P, typename OutT, typename IdxT, typename KVPair, typename ReduceOpT>
+DI void updateReducedVal(
+  int* mutex, OutT* min, KVPair* val, ReduceOpT red_op, IdxT m, IdxT gridStrideY)
+{
+  const auto lid      = threadIdx.x % raft::WarpSize;
+  const auto accrowid = threadIdx.x / P::AccThCols;
+
+  // Update each output row in order within a warp. This will resolve hang
+  // issues with pre-Volta architectures
+#pragma unroll
+  for (int j = 0; j < (raft::WarpSize / P::AccThCols); j++) {
+    if (lid == j * P::AccThCols) {
+#pragma unroll
+      for (int i = 0; i < P::AccRowsPerTh; ++i) {
+        auto rid = gridStrideY + accrowid + i * P::AccThRows;
+        if (rid < m) {
+          auto value = val[i];
+          while (atomicCAS(mutex + rid, 0, 1) == 1)
+            ;
+          __threadfence();
+          red_op(rid, min + rid, value);
+          __threadfence();
+          atomicCAS(mutex + rid, 1, 0);
+        }
+      }
+    }
+  }
+}
+
+template <typename DataT,
+          typename OutT,
+          typename IdxT,
+          typename P,
+          typename ReduceOpT,
+          typename KVPReduceOpT,
+          typename OpT,
+          typename FinalLambda>
+__launch_bounds__(P::Nthreads, 2) RAFT_KERNEL fusedL2NNkernel(OutT* min,
+                                                              const DataT* x,
+                                                              const DataT* y,
+                                                              const DataT* xn,
+                                                              const DataT* yn,
+                                                              IdxT m,
+                                                              IdxT n,
+                                                              IdxT k,
+                                                              DataT maxVal,
+                                                              int* mutex,
+                                                              ReduceOpT redOp,
+                                                              KVPReduceOpT pairRedOp,
+                                                              OpT distance_op,
+                                                              FinalLambda fin_op)
+{
+// compile only if below non-ampere arch.
+#if __CUDA_ARCH__ < 800
+  extern __shared__ char smem[];
+
+  typedef KeyValuePair<IdxT, DataT> KVPair;
+  KVPair val[P::AccRowsPerTh];
+#pragma unroll
+  for (int i = 0; i < P::AccRowsPerTh; ++i) {
+    val[i] = {0, maxVal};
+  }
+
+  // epilogue operation lambda for final value calculation
+  auto epilog_lambda = [n, pairRedOp, &val, maxVal] __device__(
+                         DataT acc[P::AccRowsPerTh][P::AccColsPerTh],
+                         DataT * regxn,
+                         DataT * regyn,
+                         IdxT gridStrideX,
+                         IdxT gridStrideY) {
+    KVPReduceOpT pairRed_op(pairRedOp);
+
+    // intra thread reduce
+    const auto acccolid = threadIdx.x % P::AccThCols;
+    const auto accrowid = threadIdx.x / P::AccThCols;
+#pragma unroll
+    for (int i = 0; i < P::AccRowsPerTh; ++i) {
+#pragma unroll
+      for (int j = 0; j < P::AccColsPerTh; ++j) {
+        auto tmpkey = acccolid + j * P::AccThCols + gridStrideX;
+        KVPair tmp  = {tmpkey, acc[i][j]};
+        if (tmpkey < n) {
+          val[i] = pairRed_op(accrowid + i * P::AccThRows + gridStrideY, tmp, val[i]);
+        }
+      }
+    }
+  };
+
+  auto rowEpilog_lambda =
+    [m, mutex, min, pairRedOp, redOp, &val, maxVal] __device__(IdxT gridStrideY) {
+      KVPReduceOpT pairRed_op(pairRedOp);
+      ReduceOpT red_op(redOp);
+
+      const auto accrowid = threadIdx.x / P::AccThCols;
+      const auto lid      = raft::laneId();
+
+    // reduce
+#pragma unroll
+      for (int i = 0; i < P::AccRowsPerTh; ++i) {
+#pragma unroll
+        for (int j = P::AccThCols / 2; j > 0; j >>= 1) {
+          // Actually, the srcLane (lid +j) should be (lid +j) % P:AccThCols,
+          // but the shfl op applies the modulo internally.
+          auto tmpkey   = raft::shfl(val[i].key, lid + j, P::AccThCols);
+          auto tmpvalue = raft::shfl(val[i].value, lid + j, P::AccThCols);
+          KVPair tmp    = {tmpkey, tmpvalue};
+          val[i]        = pairRed_op(accrowid + i * P::AccThRows + gridStrideY, tmp, val[i]);
+        }
+      }
+
+      updateReducedVal<P, OutT, IdxT, KVPair, ReduceOpT>(mutex, min, val, red_op, m, gridStrideY);
+
+    // reset the val array.
+#pragma unroll
+      for (int i = 0; i < P::AccRowsPerTh; ++i) {
+        val[i] = {0, maxVal};
+      }
+    };
+
+  IdxT lda = k, ldb = k, ldd = n;
+  constexpr bool row_major = true;
+  constexpr bool write_out = false;
+  PairwiseDistances<DataT,
+                    DataT,  // OutT (unused in PairwiseDistances)
+                    IdxT,
+                    P,
+                    decltype(distance_op),
+                    decltype(epilog_lambda),
+                    FinalLambda,
+                    decltype(rowEpilog_lambda),
+                    row_major,
+                    write_out>
+    obj(x,
+        y,
+        m,
+        n,
+        k,
+        lda,
+        ldb,
+        ldd,
+        xn,
+        yn,
+        nullptr,  // Output pointer
+        smem,
+        distance_op,
+        epilog_lambda,
+        fin_op,
+        rowEpilog_lambda);
+  obj.run();
+#endif
+}
+
+// cg::reduce functor for FusedDistanceNN used in its cutlass version
+// to output the min distance value & key(loc id).
+// This is used in fused_distance_nn/predicated_tile_iterator_reduced_vec.h
+// store_with_byte_offset() passed to cg::reduce() & select_reduce.
+template <typename AccType, typename Index, typename OutType>
+struct kvp_cg_min_reduce_op {
+  typedef typename raft::KeyValuePair<Index, AccType> KVP;
+
+  __host__ __device__ kvp_cg_min_reduce_op() noexcept {};
+
+  using AccTypeT = AccType;
+  using IndexT   = Index;
+  // functor signature.
+  __host__ __device__ KVP operator()(KVP a, KVP b) const { return a.value < b.value ? a : b; }
+
+  __host__ __device__ AccType operator()(AccType a, AccType b) const { return min(a, b); }
+
+  __host__ __device__ bool isAmin(AccType a, AccType b) const { return a < b ? true : false; }
+};
+
+template <typename DataT,
+          typename OutT,
+          typename IdxT,
+          typename Policy,
+          typename ReduceOpT,
+          typename KVPReduceOpT>
+void fusedL2NNImpl(OutT* min,
+                   const DataT* x,
+                   const DataT* y,
+                   const DataT* xn,
+                   const DataT* yn,
+                   IdxT m,
+                   IdxT n,
+                   IdxT k,
+                   int* workspace,
+                   ReduceOpT redOp,
+                   KVPReduceOpT pairRedOp,
+                   bool sqrt,
+                   bool initOutBuffer,
+                   cudaStream_t stream)
+{
+  // The kernel policy is determined by fusedL2NN.
+  typedef Policy P;
+
+  dim3 blk(P::Nthreads);
+  auto nblks            = raft::ceildiv<int>(m, P::Nthreads);
+  constexpr auto maxVal = std::numeric_limits<DataT>::max();
+  typedef KeyValuePair<IdxT, DataT> KVPair;
+
+  RAFT_CUDA_TRY(cudaMemsetAsync(workspace, 0, sizeof(int) * m, stream));
+  if (initOutBuffer) {
+    initKernel<DataT, OutT, IdxT, ReduceOpT>
+      <<<nblks, P::Nthreads, 0, stream>>>(min, m, maxVal, redOp);
+    RAFT_CUDA_TRY(cudaGetLastError());
+  }
+
+  namespace arch = raft::util::arch;
+  using AccT     = DataT;
+  ops::l2_exp_distance_op<DataT, AccT, IdxT> distance_op{sqrt};
+
+  raft::identity_op fin_op{};
+
+  auto kernel = fusedL2NNkernel<DataT,
+                                OutT,
+                                IdxT,
+                                P,
+                                ReduceOpT,
+                                KVPReduceOpT,
+                                decltype(distance_op),
+                                decltype(fin_op)>;
+
+  // Get pointer to fp32 SIMT kernel to determine the best compute architecture
+  // out of all for which the kernel was compiled for that matches closely
+  // to the current device. Other methods to determine the architecture (that do not
+  // require a pointer) can be error prone. See:
+  // https://github.com/NVIDIA/cub/issues/545
+  void* kernel_ptr   = reinterpret_cast<void*>(kernel);
+  auto runtime_arch  = arch::kernel_virtual_arch(kernel_ptr);
+  auto cutlass_range = arch::SM_range(arch::SM_80(), arch::SM_future());
+
+  if (cutlass_range.contains(runtime_arch)) {
+    // If device is SM_80 or later, use CUTLASS-based kernel.
+    using L2Op                  = raft::distance::detail::ops::l2_exp_cutlass_op<DataT, DataT>;
+    using kvp_cg_min_reduce_op_ = kvp_cg_min_reduce_op<DataT, IdxT, OutT>;
+    kvp_cg_min_reduce_op_ cg_reduce_op;
+    L2Op L2_dist_op(sqrt);
+
+    IdxT lda, ldb, ldd;
+    lda = k, ldb = k, ldd = n;
+
+    cutlassFusedDistanceNN<DataT,
+                           DataT,
+                           OutT,
+                           IdxT,
+                           P::Veclen,
+                           kvp_cg_min_reduce_op_,
+                           L2Op,
+                           ReduceOpT,
+                           KVPReduceOpT>(x,
+                                         y,
+                                         xn,
+                                         yn,
+                                         m,
+                                         n,
+                                         k,
+                                         lda,
+                                         ldb,
+                                         ldd,
+                                         min,
+                                         workspace,
+                                         cg_reduce_op,
+                                         L2_dist_op,
+                                         redOp,
+                                         pairRedOp,
+                                         stream);
+  } else {
+    // If device less than SM_80, use fp32 SIMT kernel.
+    constexpr size_t shmemSize = P::SmemSize + ((P::Mblk + P::Nblk) * sizeof(DataT));
+    dim3 grid                  = launchConfigGenerator<P>(m, n, shmemSize, kernel);
+
+    kernel<<<grid, blk, shmemSize, stream>>>(
+      min, x, y, xn, yn, m, n, k, maxVal, workspace, redOp, pairRedOp, distance_op, fin_op);
+    RAFT_CUDA_TRY(cudaGetLastError());
+  }
+}
+
+}  // namespace detail
+}  // namespace distance
+}  // namespace raft
diff --git a/cpp/include/raft/distance/detail/kernels/gram_matrix.cuh b/cpp/include/raft/distance/detail/kernels/gram_matrix.cuh
index 14b4ba12c6..4ce489bc05 100644
--- a/cpp/include/raft/distance/detail/kernels/gram_matrix.cuh
+++ b/cpp/include/raft/distance/detail/kernels/gram_matrix.cuh
@@ -22,11 +22,10 @@
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 // #include <raft/sparse/detail/cusparse_wrappers.h>
-#include <raft/sparse/distance/distance.cuh>
-#include <raft/sparse/linalg/spmm.hpp>
-
 #include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/linalg/gemm.cuh>
+#include <raft/sparse/distance/distance.cuh>
+#include <raft/sparse/linalg/spmm.hpp>
 
 namespace raft::distance::kernels::detail {
 
diff --git a/cpp/include/raft/distance/detail/kernels/kernel_factory.cuh b/cpp/include/raft/distance/detail/kernels/kernel_factory.cuh
index bb3ff1c2f5..a64f47cea4 100644
--- a/cpp/include/raft/distance/detail/kernels/kernel_factory.cuh
+++ b/cpp/include/raft/distance/detail/kernels/kernel_factory.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,6 +18,7 @@
 
 #include "gram_matrix.cuh"
 #include "kernel_matrices.cuh"
+
 #include <raft/distance/distance_types.hpp>
 #include <raft/util/cudart_utils.hpp>
 
diff --git a/cpp/include/raft/distance/detail/kernels/kernel_matrices.cuh b/cpp/include/raft/distance/detail/kernels/kernel_matrices.cuh
index 8d5b2c766e..0d64479d84 100644
--- a/cpp/include/raft/distance/detail/kernels/kernel_matrices.cuh
+++ b/cpp/include/raft/distance/detail/kernels/kernel_matrices.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,8 +17,8 @@
 #pragma once
 
 #include "gram_matrix.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/detail/kernels/rbf_fin_op.cuh>
 #include <raft/distance/distance.cuh>
 #include <raft/linalg/gemm.cuh>
diff --git a/cpp/include/raft/distance/detail/masked_nn.cuh b/cpp/include/raft/distance/detail/masked_nn.cuh
index 7bbb1ae789..3e3699766f 100644
--- a/cpp/include/raft/distance/detail/masked_nn.cuh
+++ b/cpp/include/raft/distance/detail/masked_nn.cuh
@@ -16,18 +16,20 @@
 
 #pragma once
 
-#include <limits>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/device_memory_resource.hpp>
-#include <stdint.h>
-
 #include <raft/distance/detail/compress_to_bits.cuh>
 #include <raft/distance/detail/fused_distance_nn/fused_l2_nn.cuh>
 #include <raft/distance/detail/masked_distance_base.cuh>
 #include <raft/linalg/contractions.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <stdint.h>
+
+#include <limits>
+
 namespace raft {
 namespace distance {
 namespace detail {
diff --git a/cpp/include/raft/distance/detail/pairwise_distance_cutlass_base.cuh b/cpp/include/raft/distance/detail/pairwise_distance_cutlass_base.cuh
index aeb862b06a..282097742c 100644
--- a/cpp/include/raft/distance/detail/pairwise_distance_cutlass_base.cuh
+++ b/cpp/include/raft/distance/detail/pairwise_distance_cutlass_base.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -26,23 +26,23 @@
 #define cutlass raft_cutlass
 #endif
 
+#include "pairwise_distance_epilogue_elementwise.h"
+#include "pairwise_distance_gemm.h"
+
+#include <raft/distance/detail/distance_ops/cutlass.cuh>
+#include <raft/util/cutlass_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
-#include <type_traits>
 
 #include <cutlass/cutlass.h>
 #include <cutlass/gemm/device/gemm.h>
 #include <cutlass/gemm/device/gemm_universal_adapter.h>
-
 #include <cutlass/layout/matrix.h>
 #include <cutlass/layout/tensor.h>
 #include <cutlass/matrix_coord.h>
 #include <cutlass/tensor_view.h>
 
-#include <raft/distance/detail/distance_ops/cutlass.cuh>
-#include <raft/util/cutlass_utils.cuh>
-
-#include "./pairwise_distance_epilogue_elementwise.h"
-#include "./pairwise_distance_gemm.h"
+#include <type_traits>
 
 namespace raft {
 namespace distance {
diff --git a/cpp/include/raft/distance/detail/pairwise_distance_epilogue.h b/cpp/include/raft/distance/detail/pairwise_distance_epilogue.h
index 21e7d18854..6ead09ed16 100644
--- a/cpp/include/raft/distance/detail/pairwise_distance_epilogue.h
+++ b/cpp/include/raft/distance/detail/pairwise_distance_epilogue.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -30,17 +30,16 @@ operation.
 
 #pragma once
 
+#include "./predicated_tile_iterator_normvec.h"
+
 #include <cutlass/array.h>
 #include <cutlass/cutlass.h>
-#include <cutlass/numeric_types.h>
-
-#include <cutlass/gemm/gemm.h>
-
-#include "./predicated_tile_iterator_normvec.h"
 #include <cutlass/epilogue/threadblock/default_epilogue_tensor_op.h>
 #include <cutlass/epilogue/threadblock/default_epilogue_volta_tensor_op.h>
 #include <cutlass/epilogue/threadblock/epilogue.h>
 #include <cutlass/epilogue/threadblock/epilogue_with_broadcast.h>
+#include <cutlass/gemm/gemm.h>
+#include <cutlass/numeric_types.h>
 
 ////////////////////////////////////////////////////////////////////////////////
 
diff --git a/cpp/include/raft/distance/detail/pairwise_distance_epilogue_elementwise.h b/cpp/include/raft/distance/detail/pairwise_distance_epilogue_elementwise.h
index 3e33f4d833..2b2c04b9d3 100644
--- a/cpp/include/raft/distance/detail/pairwise_distance_epilogue_elementwise.h
+++ b/cpp/include/raft/distance/detail/pairwise_distance_epilogue_elementwise.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -27,12 +27,12 @@
 
 #include <cutlass/array.h>
 #include <cutlass/cutlass.h>
+#include <cutlass/epilogue/thread/activation.h>
 #include <cutlass/functional.h>
+#include <cutlass/layout/pitch_linear.h>
 #include <cutlass/numeric_conversion.h>
 #include <cutlass/numeric_types.h>
 
-#include <cutlass/epilogue/thread/activation.h>
-
 /////////////////////////////////////////////////////////////////////////////////////////////////
 
 namespace cutlass {
diff --git a/cpp/include/raft/distance/detail/pairwise_distance_gemm.h b/cpp/include/raft/distance/detail/pairwise_distance_gemm.h
index 8dcccfc14f..aaf2689dab 100644
--- a/cpp/include/raft/distance/detail/pairwise_distance_gemm.h
+++ b/cpp/include/raft/distance/detail/pairwise_distance_gemm.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,14 @@
 
 #pragma once
 
-#include <cutlass/cutlass.h>
+#include "./pairwise_distance_epilogue.h"
 
+#include <cutlass/cutlass.h>
 #include <cutlass/gemm/kernel/default_gemm_universal.h>
 #include <cutlass/gemm/kernel/gemm_with_fused_epilogue.h>
 #include <cutlass/layout/matrix.h>
 #include <cutlass/layout/tensor.h>
 
-#include "./pairwise_distance_epilogue.h"
-
 /////////////////////////////////////////////////////////////////////////////////////////////////
 
 namespace cutlass {
diff --git a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_layout.cuh b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_layout.cuh
index f2b0e59822..95926e5c9a 100644
--- a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_layout.cuh
+++ b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_layout.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <algorithm>                                        // std::min
-#include <cstdint>                                          // size_t
 #include <raft/core/error.hpp>                              // RAFT_EXPECTS
 #include <raft/distance/detail/pairwise_matrix/params.cuh>  // pairwise_matrix_params
-#include <type_traits>                                      // std::integral_constant
+
+#include <algorithm>    // std::min
+#include <cstdint>      // size_t
+#include <type_traits>  // std::integral_constant
 namespace raft::distance::detail {
 
 /**
diff --git a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm60.cuh b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm60.cuh
index 2080fbe9cd..fc8f594275 100644
--- a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm60.cuh
+++ b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm60.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <algorithm>                                                 // std::min
 #include <raft/distance/detail/pairwise_matrix/dispatch_layout.cuh>  // dispatch_layout
 #include <raft/distance/detail/pairwise_matrix/kernel_sm60.cuh>      // pairwise_matrix_sm60_wrapper
 #include <raft/linalg/contractions.cuh>                              // raft::linalg::Policy4x4
 
+#include <algorithm>  // std::min
+
 namespace raft::distance::detail {
 
 template <typename OpT,
diff --git a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm80.cuh b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm80.cuh
index dc30f2f239..e5942653d8 100644
--- a/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm80.cuh
+++ b/cpp/include/raft/distance/detail/pairwise_matrix/dispatch_sm80.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,10 +15,11 @@
  */
 #pragma once
 
-#include <algorithm>                                                 // std::min
 #include <raft/distance/detail/pairwise_distance_cutlass_base.cuh>   // cutlassDistanceKernel
 #include <raft/distance/detail/pairwise_matrix/dispatch_layout.cuh>  // dispatch_layout
 
+#include <algorithm>  // std::min
+
 namespace raft::distance::detail {
 
 template <typename OpT,
diff --git a/cpp/include/raft/distance/detail/pairwise_matrix/kernel_sm60.cuh b/cpp/include/raft/distance/detail/pairwise_matrix/kernel_sm60.cuh
index 5393bf7389..cd9f497a06 100644
--- a/cpp/include/raft/distance/detail/pairwise_matrix/kernel_sm60.cuh
+++ b/cpp/include/raft/distance/detail/pairwise_matrix/kernel_sm60.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 #pragma once
 
-#include <cassert>                                          // assert
 #include <raft/core/operators.hpp>                          // raft::void_op
 #include <raft/distance/detail/pairwise_distance_base.cuh>  // PairwiseDistances
 #include <raft/distance/detail/pairwise_matrix/params.cuh>  // pairwise_matrix_params
 #include <raft/util/arch.cuh>                               // raft::util::arch::SM_compute_arch
 
+#include <cassert>  // assert
+
 namespace raft::distance::detail {
 
 template <typename Policy,
diff --git a/cpp/include/raft/distance/distance-ext.cuh b/cpp/include/raft/distance/distance-ext.cuh
index 7171ba605f..a634e8c995 100644
--- a/cpp/include/raft/distance/distance-ext.cuh
+++ b/cpp/include/raft/distance/distance-ext.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,7 +21,8 @@
 #include <raft/distance/detail/kernels/rbf_fin_op.cuh>  // rbf_fin_op
 #include <raft/distance/distance_types.hpp>             // raft::distance::DistanceType
 #include <raft/util/raft_explicit.hpp>                  // RAFT_EXPLICIT
-#include <rmm/device_uvector.hpp>                       // rmm::device_uvector
+
+#include <rmm/device_uvector.hpp>  // rmm::device_uvector
 
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
diff --git a/cpp/include/raft/distance/distance-inl.cuh b/cpp/include/raft/distance/distance-inl.cuh
index d17e5767b9..647c5b2908 100644
--- a/cpp/include/raft/distance/distance-inl.cuh
+++ b/cpp/include/raft/distance/distance-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,15 @@
  */
 #pragma once
 
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/detail/distance.cuh>
 #include <raft/distance/distance_types.hpp>
+
 #include <rmm/device_uvector.hpp>
-#include <type_traits>
 
-#include <raft/core/device_mdspan.hpp>
+#include <type_traits>
 
 namespace raft {
 namespace distance {
diff --git a/cpp/include/raft/distance/fused_l2_nn-ext.cuh b/cpp/include/raft/distance/fused_l2_nn-ext.cuh
index 66e9960f1d..ffc6cccab5 100644
--- a/cpp/include/raft/distance/fused_l2_nn-ext.cuh
+++ b/cpp/include/raft/distance/fused_l2_nn-ext.cuh
@@ -16,11 +16,12 @@
 
 #pragma once
 
-#include <cstdint>                                      // int64_t
-#include <raft/core/kvp.hpp>                            // raft::KeyValuePair
-#include <raft/core/resources.hpp>                      // raft::resources
+#include <raft/core/kvp.hpp>                      // raft::KeyValuePair
+#include <raft/core/resources.hpp>                // raft::resources
 #include <raft/distance/fused_distance_nn_helpers.cuh>  // include initialize and reduce operations
-#include <raft/util/raft_explicit.hpp>                  // RAFT_EXPLICIT
+#include <raft/util/raft_explicit.hpp>            // RAFT_EXPLICIT
+
+#include <cstdint>  // int64_t
 
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
diff --git a/cpp/include/raft/distance/fused_l2_nn-inl.cuh b/cpp/include/raft/distance/fused_l2_nn-inl.cuh
index 10af89e051..bf9a49d813 100644
--- a/cpp/include/raft/distance/fused_l2_nn-inl.cuh
+++ b/cpp/include/raft/distance/fused_l2_nn-inl.cuh
@@ -19,14 +19,17 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-#include <limits>
 #include <raft/core/resources.hpp>
 #include <raft/distance/detail/fused_distance_nn/fused_l2_nn.cuh>
 #include <raft/distance/fused_distance_nn_helpers.cuh>
 #include <raft/linalg/contractions.cuh>
 #include <raft/util/cuda_utils.cuh>
+
+#include <cub/cub.cuh>
+
 #include <stdint.h>
+
+#include <limits>
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/distance/kernels.cuh b/cpp/include/raft/distance/kernels.cuh
index 86f9f82406..39a6ef8c6e 100644
--- a/cpp/include/raft/distance/kernels.cuh
+++ b/cpp/include/raft/distance/kernels.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,10 +18,9 @@
 
 #include <raft/distance/detail/kernels/gram_matrix.cuh>
 #include <raft/distance/detail/kernels/kernel_factory.cuh>
-#include <raft/util/cuda_utils.cuh>
-
 #include <raft/distance/distance.cuh>
 #include <raft/linalg/gemm.cuh>
+#include <raft/util/cuda_utils.cuh>
 
 namespace raft::distance::kernels {
 
diff --git a/cpp/include/raft/distance/masked_nn.cuh b/cpp/include/raft/distance/masked_nn.cuh
index 33a6c0456d..5833a2e681 100644
--- a/cpp/include/raft/distance/masked_nn.cuh
+++ b/cpp/include/raft/distance/masked_nn.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,13 +19,15 @@
 
 #pragma once
 
-#include <limits>
 #include <raft/core/handle.hpp>
 #include <raft/distance/detail/masked_nn.cuh>
 #include <raft/distance/fused_l2_nn.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <stdint.h>
 
+#include <limits>
+
 namespace raft {
 namespace distance {
 /**
diff --git a/cpp/include/raft/label/detail/classlabels.cuh b/cpp/include/raft/label/detail/classlabels.cuh
index 6e432e050c..3742b688a0 100644
--- a/cpp/include/raft/label/detail/classlabels.cuh
+++ b/cpp/include/raft/label/detail/classlabels.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,16 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-
 #include <raft/core/operators.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
 #include <algorithm>
 
 namespace raft {
diff --git a/cpp/include/raft/label/detail/merge_labels.cuh b/cpp/include/raft/label/detail/merge_labels.cuh
index 166bb2122a..5513f16b9f 100644
--- a/cpp/include/raft/label/detail/merge_labels.cuh
+++ b/cpp/include/raft/label/detail/merge_labels.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,14 @@
 
 #pragma once
 
-#include <limits>
-#include <math.h>
-
 #include <raft/linalg/init.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <math.h>
+
+#include <limits>
+
 namespace raft {
 namespace label {
 namespace detail {
diff --git a/cpp/include/raft/linalg/add.cuh b/cpp/include/raft/linalg/add.cuh
index 26b012956a..1e903a529e 100644
--- a/cpp/include/raft/linalg/add.cuh
+++ b/cpp/include/raft/linalg/add.cuh
@@ -19,10 +19,10 @@
 #pragma once
 
 #include "detail/add.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/input_validation.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/axpy.cuh b/cpp/include/raft/linalg/axpy.cuh
index 2c901b45da..c03b30dc21 100644
--- a/cpp/include/raft/linalg/axpy.cuh
+++ b/cpp/include/raft/linalg/axpy.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,10 +19,10 @@
 #pragma once
 
 #include "detail/axpy.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 
 namespace raft::linalg {
 
diff --git a/cpp/include/raft/linalg/cholesky_r1_update.cuh b/cpp/include/raft/linalg/cholesky_r1_update.cuh
index 5c345028f2..292140b4dc 100644
--- a/cpp/include/raft/linalg/cholesky_r1_update.cuh
+++ b/cpp/include/raft/linalg/cholesky_r1_update.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/cholesky_r1_update.cuh"
+
 #include <raft/core/resource/cublas_handle.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/coalesced_reduction.cuh b/cpp/include/raft/linalg/coalesced_reduction.cuh
index afa58d73fc..a4247e618f 100644
--- a/cpp/include/raft/linalg/coalesced_reduction.cuh
+++ b/cpp/include/raft/linalg/coalesced_reduction.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,10 +19,10 @@
 #pragma once
 
 #include "detail/coalesced_reduction.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/detail/axpy.cuh b/cpp/include/raft/linalg/detail/axpy.cuh
index 8dfeab1118..0a5aaa4cb8 100644
--- a/cpp/include/raft/linalg/detail/axpy.cuh
+++ b/cpp/include/raft/linalg/detail/axpy.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <cublas_v2.h>
-#include <raft/core/resource/cublas_handle.hpp>
-
 #include "cublas_wrappers.hpp"
 
+#include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cublas_v2.h>
+
 namespace raft::linalg::detail {
 
 template <typename T, bool DevicePointerMode = false>
diff --git a/cpp/include/raft/linalg/detail/cholesky_r1_update.cuh b/cpp/include/raft/linalg/detail/cholesky_r1_update.cuh
index 34d6bf01ee..f348543cf8 100644
--- a/cpp/include/raft/linalg/detail/cholesky_r1_update.cuh
+++ b/cpp/include/raft/linalg/detail/cholesky_r1_update.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,6 +18,7 @@
 
 #include "cublas_wrappers.hpp"
 #include "cusolver_wrappers.hpp"
+
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
diff --git a/cpp/include/raft/linalg/detail/coalesced_reduction-inl.cuh b/cpp/include/raft/linalg/detail/coalesced_reduction-inl.cuh
index f3c150cbee..d580ea72c1 100644
--- a/cpp/include/raft/linalg/detail/coalesced_reduction-inl.cuh
+++ b/cpp/include/raft/linalg/detail/coalesced_reduction-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,14 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
 #include <raft/core/nvtx.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/cublas_wrappers.hpp b/cpp/include/raft/linalg/detail/cublas_wrappers.hpp
index d15e343c9a..231a209d0a 100644
--- a/cpp/include/raft/linalg/detail/cublas_wrappers.hpp
+++ b/cpp/include/raft/linalg/detail/cublas_wrappers.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <cublas_v2.h>
 #include <raft/core/cublas_macros.hpp>
 #include <raft/core/error.hpp>
 
-#include <cstdint>
 #include <cublas_v2.h>
 
+#include <cstdint>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/cublaslt_wrappers.hpp b/cpp/include/raft/linalg/detail/cublaslt_wrappers.hpp
index 423f5d29bf..ce0e45dc7a 100644
--- a/cpp/include/raft/linalg/detail/cublaslt_wrappers.hpp
+++ b/cpp/include/raft/linalg/detail/cublaslt_wrappers.hpp
@@ -24,8 +24,10 @@
 #include <raft/util/cache.hpp>
 #include <raft/util/cuda_data_type.hpp>
 
-#include <cublasLt.h>
 #include <cuda_fp16.hpp>
+
+#include <cublasLt.h>
+
 #include <type_traits>
 
 namespace raft::linalg::detail {
diff --git a/cpp/include/raft/linalg/detail/cusolver_wrappers.hpp b/cpp/include/raft/linalg/detail/cusolver_wrappers.hpp
index 79fd869083..3b056e6960 100644
--- a/cpp/include/raft/linalg/detail/cusolver_wrappers.hpp
+++ b/cpp/include/raft/linalg/detail/cusolver_wrappers.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,12 @@
 
 #pragma once
 
-#include <cusolverDn.h>
-#include <cusolverSp.h>
 #include <raft/core/cusolver_macros.hpp>
 #include <raft/util/cudart_utils.hpp>
+
+#include <cusolverDn.h>
+#include <cusolverSp.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/detail/eig.cuh b/cpp/include/raft/linalg/detail/eig.cuh
index c9f6c3c040..2a4cfd52ec 100644
--- a/cpp/include/raft/linalg/detail/eig.cuh
+++ b/cpp/include/raft/linalg/detail/eig.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,14 +17,17 @@
 #pragma once
 
 #include "cusolver_wrappers.hpp"
-#include <cuda_runtime_api.h>
+
 #include <raft/core/resource/cusolver_dn_handle.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/matrix/copy.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_runtime_api.h>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/gemv.hpp b/cpp/include/raft/linalg/detail/gemv.hpp
index c75bb87515..0c1a5235f1 100644
--- a/cpp/include/raft/linalg/detail/gemv.hpp
+++ b/cpp/include/raft/linalg/detail/gemv.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <cublas_v2.h>
-#include <raft/core/resource/cublas_handle.hpp>
-
 #include "cublas_wrappers.hpp"
 
+#include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resources.hpp>
 
+#include <cublas_v2.h>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/lanczos.cuh b/cpp/include/raft/linalg/detail/lanczos.cuh
index 3ab020bfd4..134ef3ef36 100644
--- a/cpp/include/raft/linalg/detail/lanczos.cuh
+++ b/cpp/include/raft/linalg/detail/lanczos.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,21 +19,23 @@
 // for cmath:
 #define _USE_MATH_DEFINES
 
-#include <cmath>
+#include "cublas_wrappers.hpp"
+
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
-
-#include <cuda.h>
-#include <curand.h>
-
-#include "cublas_wrappers.hpp"
 #include <raft/core/resources.hpp>
 #include <raft/spectral/detail/lapack.hpp>
 #include <raft/spectral/detail/warn_dbg.hpp>
 #include <raft/spectral/matrix_wrappers.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda.h>
+
+#include <curand.h>
+
+#include <cmath>
+#include <vector>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/lstsq.cuh b/cpp/include/raft/linalg/detail/lstsq.cuh
index 128757d1d8..c2842fda30 100644
--- a/cpp/include/raft/linalg/detail/lstsq.cuh
+++ b/cpp/include/raft/linalg/detail/lstsq.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include <common/nvtx.hpp>
+
 #include <raft/common/nvtx.hpp>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream_pool.hpp>
@@ -32,6 +33,7 @@
 #include <raft/linalg/transpose.cuh>
 #include <raft/matrix/math.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
diff --git a/cpp/include/raft/linalg/detail/map_then_reduce.cuh b/cpp/include/raft/linalg/detail/map_then_reduce.cuh
index d1e211f8d2..9ea100108f 100644
--- a/cpp/include/raft/linalg/detail/map_then_reduce.cuh
+++ b/cpp/include/raft/linalg/detail/map_then_reduce.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,12 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
 #include <raft/core/resources.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/vectorized.cuh>
 
+#include <cub/cub.cuh>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/qr.cuh b/cpp/include/raft/linalg/detail/qr.cuh
index 16a721dfd3..6233786812 100644
--- a/cpp/include/raft/linalg/detail/qr.cuh
+++ b/cpp/include/raft/linalg/detail/qr.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,9 +18,11 @@
 
 #include "cublas_wrappers.hpp"
 #include "cusolver_wrappers.hpp"
+
 #include <raft/core/resource/cusolver_dn_handle.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/matrix/triangular.cuh>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/include/raft/linalg/detail/reduce_cols_by_key.cuh b/cpp/include/raft/linalg/detail/reduce_cols_by_key.cuh
index b726e3ea5a..4d6b82920b 100644
--- a/cpp/include/raft/linalg/detail/reduce_cols_by_key.cuh
+++ b/cpp/include/raft/linalg/detail/reduce_cols_by_key.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,14 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-#include <limits>
 #include <raft/util/cuda_utils.cuh>
+
+#include <cub/cub.cuh>
+
 #include <stdlib.h>
 
+#include <limits>
+
 namespace raft {
 namespace linalg {
 namespace detail {
diff --git a/cpp/include/raft/linalg/detail/strided_reduction.cuh b/cpp/include/raft/linalg/detail/strided_reduction.cuh
index aef346bd4b..617ac6d874 100644
--- a/cpp/include/raft/linalg/detail/strided_reduction.cuh
+++ b/cpp/include/raft/linalg/detail/strided_reduction.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,12 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
 #include <raft/core/operators.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
+
+#include <cub/cub.cuh>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/detail/svd.cuh b/cpp/include/raft/linalg/detail/svd.cuh
index 5a4851bf6e..79b64c7447 100644
--- a/cpp/include/raft/linalg/detail/svd.cuh
+++ b/cpp/include/raft/linalg/detail/svd.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,21 +18,22 @@
 
 #include "cublas_wrappers.hpp"
 #include "cusolver_wrappers.hpp"
+
+#include <raft/common/nvtx.hpp>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cusolver_dn_handle.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/linalg/eig.cuh>
 #include <raft/linalg/gemm.cuh>
 #include <raft/linalg/transpose.cuh>
-
-#include <raft/common/nvtx.hpp>
-#include <raft/core/resources.hpp>
 #include <raft/matrix/diagonal.cuh>
 #include <raft/matrix/math.cuh>
 #include <raft/matrix/norm.cuh>
 #include <raft/matrix/reverse.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/include/raft/linalg/detail/transpose.cuh b/cpp/include/raft/linalg/detail/transpose.cuh
index bbd71a4cf1..999e7f1974 100644
--- a/cpp/include/raft/linalg/detail/transpose.cuh
+++ b/cpp/include/raft/linalg/detail/transpose.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -22,7 +22,9 @@
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+
 #include <rmm/exec_policy.hpp>
+
 #include <thrust/for_each.h>
 #include <thrust/iterator/counting_iterator.h>
 
diff --git a/cpp/include/raft/linalg/divide.cuh b/cpp/include/raft/linalg/divide.cuh
index 17ec5c3136..d23c7d60a6 100644
--- a/cpp/include/raft/linalg/divide.cuh
+++ b/cpp/include/raft/linalg/divide.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,9 +19,9 @@
 #pragma once
 
 #include "detail/divide.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/input_validation.hpp>
 
diff --git a/cpp/include/raft/linalg/dot.cuh b/cpp/include/raft/linalg/dot.cuh
index 9db9074c35..5a2ee75c4e 100644
--- a/cpp/include/raft/linalg/dot.cuh
+++ b/cpp/include/raft/linalg/dot.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,13 +18,12 @@
 
 #pragma once
 
-#include <raft/core/resource/cublas_handle.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cublas_handle.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
 
 namespace raft::linalg {
 
diff --git a/cpp/include/raft/linalg/eig.cuh b/cpp/include/raft/linalg/eig.cuh
index 57f3b61388..9f03f54f9a 100644
--- a/cpp/include/raft/linalg/eig.cuh
+++ b/cpp/include/raft/linalg/eig.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,9 +19,9 @@
 #pragma once
 
 #include "detail/eig.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 
 namespace raft {
 namespace linalg {
diff --git a/cpp/include/raft/linalg/gemv.cuh b/cpp/include/raft/linalg/gemv.cuh
index 610ea07f96..31bad62930 100644
--- a/cpp/include/raft/linalg/gemv.cuh
+++ b/cpp/include/raft/linalg/gemv.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,12 +19,12 @@
 #pragma once
 
 #include "detail/gemv.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/input_validation.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/map_reduce.cuh b/cpp/include/raft/linalg/map_reduce.cuh
index f17caa478b..1886c941b9 100644
--- a/cpp/include/raft/linalg/map_reduce.cuh
+++ b/cpp/include/raft/linalg/map_reduce.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,9 +19,9 @@
 #pragma once
 
 #include "detail/map_then_reduce.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 
 namespace raft::linalg {
 
diff --git a/cpp/include/raft/linalg/matrix_vector_op.cuh b/cpp/include/raft/linalg/matrix_vector_op.cuh
index a65f6ed390..1f6dfd785b 100644
--- a/cpp/include/raft/linalg/matrix_vector_op.cuh
+++ b/cpp/include/raft/linalg/matrix_vector_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,9 +20,9 @@
 
 #include "detail/matrix_vector_op.cuh"
 #include "linalg_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/input_validation.hpp>
 
diff --git a/cpp/include/raft/linalg/mean_squared_error.cuh b/cpp/include/raft/linalg/mean_squared_error.cuh
index b59a0fcef7..9e753d3a31 100644
--- a/cpp/include/raft/linalg/mean_squared_error.cuh
+++ b/cpp/include/raft/linalg/mean_squared_error.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/mean_squared_error.cuh"
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 
diff --git a/cpp/include/raft/linalg/multiply.cuh b/cpp/include/raft/linalg/multiply.cuh
index 9973a3cc6c..1a7668f8f2 100644
--- a/cpp/include/raft/linalg/multiply.cuh
+++ b/cpp/include/raft/linalg/multiply.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,10 +19,10 @@
 #pragma once
 
 #include "detail/multiply.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/input_validation.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/norm.cuh b/cpp/include/raft/linalg/norm.cuh
index 0d472c5476..97a5d6135d 100644
--- a/cpp/include/raft/linalg/norm.cuh
+++ b/cpp/include/raft/linalg/norm.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,11 +20,11 @@
 
 #include "detail/norm.cuh"
 #include "linalg_types.hpp"
-#include <raft/core/mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/mdspan.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/norm_types.hpp>
 #include <raft/util/input_validation.hpp>
 
diff --git a/cpp/include/raft/linalg/normalize.cuh b/cpp/include/raft/linalg/normalize.cuh
index 6169cf1dfa..1f60860c8c 100644
--- a/cpp/include/raft/linalg/normalize.cuh
+++ b/cpp/include/raft/linalg/normalize.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,9 +17,9 @@
 #pragma once
 
 #include "detail/normalize.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/norm_types.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/qr.cuh b/cpp/include/raft/linalg/qr.cuh
index 022c382e67..b032cbfa3a 100644
--- a/cpp/include/raft/linalg/qr.cuh
+++ b/cpp/include/raft/linalg/qr.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/qr.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
diff --git a/cpp/include/raft/linalg/reduce.cuh b/cpp/include/raft/linalg/reduce.cuh
index 3181dd0224..a4523c6926 100644
--- a/cpp/include/raft/linalg/reduce.cuh
+++ b/cpp/include/raft/linalg/reduce.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,10 +20,10 @@
 
 #include "detail/reduce.cuh"
 #include "linalg_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/input_validation.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/reduce_cols_by_key.cuh b/cpp/include/raft/linalg/reduce_cols_by_key.cuh
index 5ed0fb7407..77ad8a9a80 100644
--- a/cpp/include/raft/linalg/reduce_cols_by_key.cuh
+++ b/cpp/include/raft/linalg/reduce_cols_by_key.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,9 +19,9 @@
 #pragma once
 
 #include "detail/reduce_cols_by_key.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/reduce_rows_by_key.cuh b/cpp/include/raft/linalg/reduce_rows_by_key.cuh
index 7d93c3946f..2bb14729f4 100644
--- a/cpp/include/raft/linalg/reduce_rows_by_key.cuh
+++ b/cpp/include/raft/linalg/reduce_rows_by_key.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,9 +19,9 @@
 #pragma once
 
 #include "detail/reduce_rows_by_key.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/rsvd.cuh b/cpp/include/raft/linalg/rsvd.cuh
index 163f360481..a90ba165ed 100644
--- a/cpp/include/raft/linalg/rsvd.cuh
+++ b/cpp/include/raft/linalg/rsvd.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/rsvd.cuh"
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 
diff --git a/cpp/include/raft/linalg/strided_reduction.cuh b/cpp/include/raft/linalg/strided_reduction.cuh
index c7ff000e00..ac97c3cd68 100644
--- a/cpp/include/raft/linalg/strided_reduction.cuh
+++ b/cpp/include/raft/linalg/strided_reduction.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,10 +20,10 @@
 #pragma once
 
 #include "detail/strided_reduction.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
 #include <type_traits>
diff --git a/cpp/include/raft/linalg/subtract.cuh b/cpp/include/raft/linalg/subtract.cuh
index f4243f9dc1..e7b5c6b65a 100644
--- a/cpp/include/raft/linalg/subtract.cuh
+++ b/cpp/include/raft/linalg/subtract.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,10 +20,10 @@
 #pragma once
 
 #include "detail/subtract.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/input_validation.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/linalg/svd.cuh b/cpp/include/raft/linalg/svd.cuh
index f7071de75b..1c57515a47 100644
--- a/cpp/include/raft/linalg/svd.cuh
+++ b/cpp/include/raft/linalg/svd.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/svd.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 
 #include <optional>
diff --git a/cpp/include/raft/linalg/transpose.cuh b/cpp/include/raft/linalg/transpose.cuh
index 1b46082fbe..6f693d0e13 100644
--- a/cpp/include/raft/linalg/transpose.cuh
+++ b/cpp/include/raft/linalg/transpose.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #pragma once
 
 #include "detail/transpose.cuh"
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/resources.hpp>
 
diff --git a/cpp/include/raft/matrix/detail/columnWiseSort.cuh b/cpp/include/raft/matrix/detail/columnWiseSort.cuh
index 652c4fda0f..a02621d054 100644
--- a/cpp/include/raft/matrix/detail/columnWiseSort.cuh
+++ b/cpp/include/raft/matrix/detail/columnWiseSort.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,13 @@
 
 #pragma once
 
-#include <cstddef>
+#include <raft/util/cuda_utils.cuh>
+
 #include <cub/cub.cuh>
+
+#include <cstddef>
 #include <limits>
 #include <map>
-#include <raft/util/cuda_utils.cuh>
 
 #define INST_BLOCK_SORT(keyIn, keyOut, valueInOut, rows, columns, blockSize, elemPT, stream)     \
   devKeyValSortColumnPerRow<InType, OutType, blockSize, elemPT><<<rows, blockSize, 0, stream>>>( \
diff --git a/cpp/include/raft/matrix/detail/gather.cuh b/cpp/include/raft/matrix/detail/gather.cuh
index da5fd8c103..651fec81c3 100644
--- a/cpp/include/raft/matrix/detail/gather.cuh
+++ b/cpp/include/raft/matrix/detail/gather.cuh
@@ -16,10 +16,11 @@
 
 #pragma once
 
-#include <functional>
 #include <raft/core/operators.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <functional>
+
 namespace raft {
 namespace matrix {
 namespace detail {
diff --git a/cpp/include/raft/matrix/detail/gather_inplace.cuh b/cpp/include/raft/matrix/detail/gather_inplace.cuh
index cc510e068b..a37ba550f9 100644
--- a/cpp/include/raft/matrix/detail/gather_inplace.cuh
+++ b/cpp/include/raft/matrix/detail/gather_inplace.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/linalg/map.cuh>
 #include <raft/util/fast_int_div.cuh>
+
 #include <thrust/iterator/counting_iterator.h>
 
 namespace raft {
diff --git a/cpp/include/raft/matrix/detail/linewise_op.cuh b/cpp/include/raft/matrix/detail/linewise_op.cuh
index 6061fe6aee..4a2e45b161 100644
--- a/cpp/include/raft/matrix/detail/linewise_op.cuh
+++ b/cpp/include/raft/matrix/detail/linewise_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -22,9 +22,10 @@
 #include <raft/util/pow2_utils.cuh>
 #include <raft/util/vectorized.cuh>
 
-#include <algorithm>
 #include <thrust/tuple.h>
 
+#include <algorithm>
+
 namespace raft {
 namespace matrix {
 namespace detail {
diff --git a/cpp/include/raft/matrix/detail/math.cuh b/cpp/include/raft/matrix/detail/math.cuh
index 9e9d7f8b3b..e1b6b2ef71 100644
--- a/cpp/include/raft/matrix/detail/math.cuh
+++ b/cpp/include/raft/matrix/detail/math.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,18 +16,19 @@
 
 #pragma once
 
-#include <raft/core/resources.hpp>
-
-#include <cub/cub.cuh>
 #include <raft/core/operators.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/linalg/binary_op.cuh>
 #include <raft/linalg/map_then_reduce.cuh>
 #include <raft/linalg/matrix_vector_op.cuh>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
 namespace raft {
 namespace matrix {
 namespace detail {
diff --git a/cpp/include/raft/matrix/detail/matrix.cuh b/cpp/include/raft/matrix/detail/matrix.cuh
index 2fa741fd96..6ed9ffa645 100644
--- a/cpp/include/raft/matrix/detail/matrix.cuh
+++ b/cpp/include/raft/matrix/detail/matrix.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,21 +17,22 @@
 #pragma once
 
 #include <raft/core/resource/cublas_handle.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/util/cache_util.cuh>
 #include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
 
 #include <rmm/exec_policy.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/for_each.h>
 #include <thrust/iterator/counting_iterator.h>
 
+#include <cusolverDn.h>
+
 #include <algorithm>
 #include <cstddef>
-#include <cuda_runtime.h>
-#include <cusolverDn.h>
-#include <raft/core/resources.hpp>
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-#include <raft/util/cudart_utils.hpp>
 
 namespace raft {
 namespace matrix {
diff --git a/cpp/include/raft/matrix/detail/print.hpp b/cpp/include/raft/matrix/detail/print.hpp
index 0b93819b97..92881d5b12 100644
--- a/cpp/include/raft/matrix/detail/print.hpp
+++ b/cpp/include/raft/matrix/detail/print.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,21 +16,22 @@
 
 #pragma once
 
+#include <raft/core/resources.hpp>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/util/cache_util.cuh>
 #include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
 
 #include <rmm/exec_policy.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/for_each.h>
 #include <thrust/iterator/counting_iterator.h>
 
+#include <cusolverDn.h>
+
 #include <algorithm>
 #include <cstddef>
-#include <cuda_runtime.h>
-#include <cusolverDn.h>
-#include <raft/core/resources.hpp>
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-#include <raft/util/cudart_utils.hpp>
 
 namespace raft::matrix::detail {
 
diff --git a/cpp/include/raft/matrix/detail/scatter_inplace.cuh b/cpp/include/raft/matrix/detail/scatter_inplace.cuh
index 3a57c5478b..6a60e5006b 100644
--- a/cpp/include/raft/matrix/detail/scatter_inplace.cuh
+++ b/cpp/include/raft/matrix/detail/scatter_inplace.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,16 @@
  */
 #pragma once
 
-#include <cstdint>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/linalg/map.cuh>
 #include <raft/util/cuda_dev_essentials.cuh>
 #include <raft/util/fast_int_div.cuh>
+
 #include <thrust/iterator/counting_iterator.h>
 
+#include <cstdint>
+
 namespace raft {
 namespace matrix {
 namespace detail {
diff --git a/cpp/include/raft/matrix/detail/select_k-ext.cuh b/cpp/include/raft/matrix/detail/select_k-ext.cuh
index e8db6827b5..6a7847d8a0 100644
--- a/cpp/include/raft/matrix/detail/select_k-ext.cuh
+++ b/cpp/include/raft/matrix/detail/select_k-ext.cuh
@@ -16,14 +16,17 @@
 
 #pragma once
 
-#include <cstdint>      // uint32_t
-#include <cuda_fp16.h>  // __half
 #include <raft/core/device_resources.hpp>
 #include <raft/matrix/select_k_types.hpp>
-#include <raft/util/raft_explicit.hpp>               // RAFT_EXPLICIT
+#include <raft/util/raft_explicit.hpp>  // RAFT_EXPLICIT
+
 #include <rmm/cuda_stream_view.hpp>                  // rmm:cuda_stream_view
 #include <rmm/mr/device/device_memory_resource.hpp>  // rmm::mr::device_memory_resource
 
+#include <cuda_fp16.h>  // __half
+
+#include <cstdint>  // uint32_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::matrix::detail {
diff --git a/cpp/include/raft/matrix/detail/select_radix.cuh b/cpp/include/raft/matrix/detail/select_radix.cuh
index 16b9ac0c6d..82983b7cd2 100644
--- a/cpp/include/raft/matrix/detail/select_radix.cuh
+++ b/cpp/include/raft/matrix/detail/select_radix.cuh
@@ -29,15 +29,15 @@
 #include <raft/util/pow2_utils.cuh>
 #include <raft/util/vectorized.cuh>
 
+#include <rmm/device_uvector.hpp>
+#include <rmm/mr/device/device_memory_resource.hpp>
+#include <rmm/mr/device/managed_memory_resource.hpp>
+
 #include <cub/block/block_load.cuh>
 #include <cub/block/block_scan.cuh>
 #include <cub/block/block_store.cuh>
 #include <cub/block/radix_rank_sort_operations.cuh>
 
-#include <rmm/device_uvector.hpp>
-#include <rmm/mr/device/device_memory_resource.hpp>
-#include <rmm/mr/device/managed_memory_resource.hpp>
-
 namespace raft::matrix::detail::select::radix {
 namespace impl {
 
diff --git a/cpp/include/raft/matrix/detail/select_warpsort.cuh b/cpp/include/raft/matrix/detail/select_warpsort.cuh
index 7cd43b030b..572558153d 100644
--- a/cpp/include/raft/matrix/detail/select_warpsort.cuh
+++ b/cpp/include/raft/matrix/detail/select_warpsort.cuh
@@ -27,13 +27,13 @@
 #include <raft/util/integer_utils.hpp>
 #include <raft/util/pow2_utils.cuh>
 
+#include <rmm/device_uvector.hpp>
+#include <rmm/mr/device/device_memory_resource.hpp>
+
 #include <algorithm>
 #include <functional>
 #include <type_traits>
 
-#include <rmm/device_uvector.hpp>
-#include <rmm/mr/device/device_memory_resource.hpp>
-
 /*
   Three APIs of different scopes are provided:
     1. host function: select_k()
diff --git a/cpp/include/raft/matrix/matrix.cuh b/cpp/include/raft/matrix/matrix.cuh
index 63c33ff034..29cc6fb75b 100644
--- a/cpp/include/raft/matrix/matrix.cuh
+++ b/cpp/include/raft/matrix/matrix.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -30,10 +30,10 @@
 
 #include "detail/linewise_op.cuh"
 #include "detail/matrix.cuh"
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/common/nvtx.hpp>
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 
 namespace raft {
 namespace matrix {
diff --git a/cpp/include/raft/matrix/select_k.cuh b/cpp/include/raft/matrix/select_k.cuh
index e2d94c67ae..2efa146495 100644
--- a/cpp/include/raft/matrix/select_k.cuh
+++ b/cpp/include/raft/matrix/select_k.cuh
@@ -17,10 +17,10 @@
 #pragma once
 
 #include "detail/select_k.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/nvtx.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/matrix/select_k_types.hpp>
 
diff --git a/cpp/include/raft/neighbors/ball_cover-ext.cuh b/cpp/include/raft/neighbors/ball_cover-ext.cuh
index 3d0b3c7858..2a83a57749 100644
--- a/cpp/include/raft/neighbors/ball_cover-ext.cuh
+++ b/cpp/include/raft/neighbors/ball_cover-ext.cuh
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <cstdint>                              // uint32_t
 #include <raft/distance/distance_types.hpp>     // raft::distance::DistanceType
 #include <raft/neighbors/ball_cover_types.hpp>  // BallCoverIndex
 #include <raft/util/raft_explicit.hpp>          // RAFT_EXPLICIT
 
+#include <cstdint>  // uint32_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::ball_cover {
diff --git a/cpp/include/raft/neighbors/ball_cover-inl.cuh b/cpp/include/raft/neighbors/ball_cover-inl.cuh
index cdf7c30e89..4af2a6772c 100644
--- a/cpp/include/raft/neighbors/ball_cover-inl.cuh
+++ b/cpp/include/raft/neighbors/ball_cover-inl.cuh
@@ -18,14 +18,15 @@
 
 #pragma once
 
-#include <cstdint>
-
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/ball_cover_types.hpp>
 #include <raft/spatial/knn/detail/ball_cover.cuh>
 #include <raft/spatial/knn/detail/ball_cover/common.cuh>
+
 #include <thrust/transform.h>
 
+#include <cstdint>
+
 namespace raft::neighbors::ball_cover {
 
 /**
diff --git a/cpp/include/raft/neighbors/ball_cover_types.hpp b/cpp/include/raft/neighbors/ball_cover_types.hpp
index dc96f0d45b..a627a1d234 100644
--- a/cpp/include/raft/neighbors/ball_cover_types.hpp
+++ b/cpp/include/raft/neighbors/ball_cover_types.hpp
@@ -16,14 +16,16 @@
 
 #pragma once
 
-#include <cstdint>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cstdint>
+
 namespace raft::neighbors::ball_cover {
 
 /**
diff --git a/cpp/include/raft/neighbors/brute_force-ext.cuh b/cpp/include/raft/neighbors/brute_force-ext.cuh
index e8c25f355b..4055c253c8 100644
--- a/cpp/include/raft/neighbors/brute_force-ext.cuh
+++ b/cpp/include/raft/neighbors/brute_force-ext.cuh
@@ -16,8 +16,6 @@
 
 #pragma once
 
-#include <optional>
-
 #include <raft/core/device_mdspan.hpp>       // raft::device_matrix_view
 #include <raft/core/operators.hpp>           // raft::identity_op
 #include <raft/core/resources.hpp>           // raft::resources
@@ -25,6 +23,8 @@
 #include <raft/neighbors/brute_force_types.hpp>
 #include <raft/util/raft_explicit.hpp>  // RAFT_EXPLICIT
 
+#include <optional>
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::brute_force {
diff --git a/cpp/include/raft/neighbors/brute_force_types.hpp b/cpp/include/raft/neighbors/brute_force_types.hpp
index 176b32a866..a8f073edc6 100644
--- a/cpp/include/raft/neighbors/brute_force_types.hpp
+++ b/cpp/include/raft/neighbors/brute_force_types.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,19 +17,18 @@
 #pragma once
 
 #include "ann_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/copy.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
+#include <raft/core/logger.hpp>
 #include <raft/core/mdspan_types.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/neighbors_types.hpp>
 
-#include <raft/core/logger.hpp>
-
 namespace raft::neighbors::brute_force {
 /**
  * @addtogroup brute_force_knn
diff --git a/cpp/include/raft/neighbors/cagra.cuh b/cpp/include/raft/neighbors/cagra.cuh
index 384ed05e1f..b8258297e6 100644
--- a/cpp/include/raft/neighbors/cagra.cuh
+++ b/cpp/include/raft/neighbors/cagra.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -25,6 +25,7 @@
 #include <raft/core/mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/neighbors/cagra_types.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 
 namespace raft::neighbors::cagra {
diff --git a/cpp/include/raft/neighbors/cagra_types.hpp b/cpp/include/raft/neighbors/cagra_types.hpp
index cc76d70fbf..0f574ae5bb 100644
--- a/cpp/include/raft/neighbors/cagra_types.hpp
+++ b/cpp/include/raft/neighbors/cagra_types.hpp
@@ -17,24 +17,24 @@
 #pragma once
 
 #include "ann_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
+#include <raft/core/logger.hpp>
 #include <raft/core/mdspan_types.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/detail/cagra/utils.hpp>
 #include <raft/util/integer_utils.hpp>
 
+#include <thrust/fill.h>
+
 #include <memory>
 #include <optional>
 #include <string>
-#include <thrust/fill.h>
 #include <type_traits>
-
-#include <raft/core/logger.hpp>
 namespace raft::neighbors::cagra {
 /**
  * @addtogroup cagra
diff --git a/cpp/include/raft/neighbors/detail/cagra/bitonic.hpp b/cpp/include/raft/neighbors/detail/cagra/bitonic.hpp
index 9fca7f8ebd..1d4c77af6f 100644
--- a/cpp/include/raft/neighbors/detail/cagra/bitonic.hpp
+++ b/cpp/include/raft/neighbors/detail/cagra/bitonic.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,10 @@
  */
 #pragma once
 
-#include <cstdint>
 #include <raft/core/detail/macros.hpp>
 
+#include <cstdint>
+
 namespace raft::neighbors::cagra::detail {
 namespace bitonic {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/cagra_build.cuh b/cpp/include/raft/neighbors/detail/cagra/cagra_build.cuh
old mode 100755
new mode 100644
index e7f7d96a13..08cc2beaeb
--- a/cpp/include/raft/neighbors/detail/cagra/cagra_build.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/cagra_build.cuh
@@ -17,10 +17,6 @@
 
 #include "../../cagra_types.hpp"
 #include "graph_core.cuh"
-#include <chrono>
-#include <cstdio>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
 
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
@@ -28,14 +24,18 @@
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
 #include <raft/neighbors/detail/refine.cuh>
 #include <raft/neighbors/ivf_pq.cuh>
 #include <raft/neighbors/ivf_pq_types.hpp>
 #include <raft/neighbors/nn_descent.cuh>
 #include <raft/neighbors/refine.cuh>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
+
+#include <chrono>
+#include <cstdio>
+#include <vector>
 
 namespace raft::neighbors::cagra::detail {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/cagra_search.cuh b/cpp/include/raft/neighbors/detail/cagra/cagra_search.cuh
index 40cc7c76fb..0832e75633 100644
--- a/cpp/include/raft/neighbors/detail/cagra/cagra_search.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/cagra_search.cuh
@@ -16,21 +16,22 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/neighbors/detail/ivf_pq_search.cuh>
-#include <raft/neighbors/sample_filter_types.hpp>
-#include <raft/spatial/knn/detail/ann_utils.cuh>
+#include "factory.cuh"
+#include "search_plan.cuh"
+#include "search_single_cta.cuh"
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/nvtx.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/neighbors/cagra_types.hpp>
-#include <rmm/cuda_stream_view.hpp>
+#include <raft/neighbors/detail/ivf_common.cuh>
+#include <raft/neighbors/detail/ivf_pq_search.cuh>
+#include <raft/neighbors/sample_filter_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 
-#include "factory.cuh"
-#include "search_plan.cuh"
-#include "search_single_cta.cuh"
+#include <rmm/cuda_stream_view.hpp>
 
 namespace raft::neighbors::cagra::detail {
 
@@ -181,13 +182,14 @@ void search_main(raft::resources const& res,
   // and divide the values by kDivisor. Here we restore the original scale.
   constexpr float kScale = spatial::knn::detail::utils::config<T>::kDivisor /
                            spatial::knn::detail::utils::config<DistanceT>::kDivisor;
-  ivf_pq::detail::postprocess_distances(dist_out,
-                                        dist_in,
-                                        index.metric(),
-                                        distances.extent(0),
-                                        distances.extent(1),
-                                        kScale,
-                                        resource::get_cuda_stream(res));
+  ivf::detail::postprocess_distances(dist_out,
+                                     dist_in,
+                                     index.metric(),
+                                     distances.extent(0),
+                                     distances.extent(1),
+                                     kScale,
+                                     true,
+                                     resource::get_cuda_stream(res));
 }
 /** @} */  // end group cagra
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/cagra_serialize.cuh b/cpp/include/raft/neighbors/detail/cagra/cagra_serialize.cuh
index 84a9ce96cd..d7bd27222b 100644
--- a/cpp/include/raft/neighbors/detail/cagra/cagra_serialize.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/cagra_serialize.cuh
@@ -16,8 +16,6 @@
 
 #pragma once
 
-#include <cstddef>
-#include <cstdint>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/mdspan_types.hpp>
@@ -26,6 +24,8 @@
 #include <raft/core/serialize.hpp>
 #include <raft/neighbors/cagra_types.hpp>
 
+#include <cstddef>
+#include <cstdint>
 #include <fstream>
 #include <type_traits>
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/compute_distance.hpp b/cpp/include/raft/neighbors/detail/cagra/compute_distance.hpp
index 7f2e8b34cb..3732dcf3fe 100644
--- a/cpp/include/raft/neighbors/detail/cagra/compute_distance.hpp
+++ b/cpp/include/raft/neighbors/detail/cagra/compute_distance.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
 #include "device_common.hpp"
 #include "hashmap.hpp"
 #include "utils.hpp"
+
+#include <raft/spatial/knn/detail/ann_utils.cuh>
+
 #include <type_traits>
 
 namespace raft::neighbors::cagra::detail {
diff --git a/cpp/include/raft/neighbors/detail/cagra/device_common.hpp b/cpp/include/raft/neighbors/detail/cagra/device_common.hpp
index b1a2207a4e..cd7469b55e 100644
--- a/cpp/include/raft/neighbors/detail/cagra/device_common.hpp
+++ b/cpp/include/raft/neighbors/detail/cagra/device_common.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,13 @@
 #pragma once
 
 #include "utils.hpp"
+
+#include <raft/core/detail/macros.hpp>
+
+#include <cuda_fp16.h>
+
 #include <cfloat>
 #include <cstdint>
-#include <cuda_fp16.h>
-#include <raft/core/detail/macros.hpp>
 
 namespace raft::neighbors::cagra::detail {
 namespace device {
diff --git a/cpp/include/raft/neighbors/detail/cagra/factory.cuh b/cpp/include/raft/neighbors/detail/cagra/factory.cuh
index 0002dd8b2a..0aee912e25 100644
--- a/cpp/include/raft/neighbors/detail/cagra/factory.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/factory.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,6 +20,7 @@
 #include "search_multi_kernel.cuh"
 #include "search_plan.cuh"
 #include "search_single_cta.cuh"
+
 #include <raft/neighbors/sample_filter_types.hpp>
 
 namespace raft::neighbors::cagra::detail {
diff --git a/cpp/include/raft/neighbors/detail/cagra/graph_core.cuh b/cpp/include/raft/neighbors/detail/cagra/graph_core.cuh
index 192850cf2d..e4e3ea3512 100644
--- a/cpp/include/raft/neighbors/detail/cagra/graph_core.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/graph_core.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,26 +15,28 @@
  */
 #pragma once
 
-#include <cassert>
-#include <climits>
-#include <cuda_fp16.h>
-#include <float.h>
-#include <iostream>
-#include <memory>
-#include <omp.h>
+#include "utils.hpp"
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/spatial/knn/detail/ann_utils.cuh>
-#include <random>
-#include <sys/time.h>
-
 #include <raft/util/bitonic_sort.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 
-#include "utils.hpp"
+#include <cuda_fp16.h>
+
+#include <float.h>
+#include <omp.h>
+#include <sys/time.h>
+
+#include <cassert>
+#include <climits>
+#include <iostream>
+#include <memory>
+#include <random>
 
 namespace raft::neighbors::cagra::detail {
 namespace graph {
diff --git a/cpp/include/raft/neighbors/detail/cagra/hashmap.hpp b/cpp/include/raft/neighbors/detail/cagra/hashmap.hpp
index ed4763e475..034bca6178 100644
--- a/cpp/include/raft/neighbors/detail/cagra/hashmap.hpp
+++ b/cpp/include/raft/neighbors/detail/cagra/hashmap.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,12 @@
 #pragma once
 
 #include "utils.hpp"
-#include <cstdint>
+
 #include <raft/core/detail/macros.hpp>
 #include <raft/util/device_atomics.cuh>
 
+#include <cstdint>
+
 // #pragma GCC diagnostic push
 // #pragma GCC diagnostic ignored
 // #pragma GCC diagnostic pop
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_multi_cta.cuh b/cpp/include/raft/neighbors/detail/cagra/search_multi_cta.cuh
index 010b0a6f80..1fcd159959 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_multi_cta.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_multi_cta.cuh
@@ -15,20 +15,6 @@
  */
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <algorithm>
-#include <cassert>
-#include <iostream>
-#include <memory>
-#include <numeric>
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_properties.hpp>
-#include <raft/core/resources.hpp>
-
-#include <vector>
-
 #include "bitonic.hpp"
 #include "compute_distance.hpp"
 #include "device_common.hpp"
@@ -37,10 +23,23 @@
 #include "search_plan.cuh"
 #include "topk_for_cagra/topk_core.cuh"  // TODO replace with raft topk if possible
 #include "utils.hpp"
+
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_properties.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 #include <raft/util/cudart_utils.hpp>  // RAFT_CUDA_TRY_NOT_THROW is used TODO(tfeher): consider moving this to cuda_rt_essentials.hpp
 
+#include <algorithm>
+#include <cassert>
+#include <iostream>
+#include <memory>
+#include <numeric>
+#include <vector>
+
 namespace raft::neighbors::cagra::detail {
 namespace multi_cta_search {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_multi_cta_kernel-inl.cuh b/cpp/include/raft/neighbors/detail/cagra/search_multi_cta_kernel-inl.cuh
index 8b394befd7..30f56780d6 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_multi_cta_kernel-inl.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_multi_cta_kernel-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,21 +15,6 @@
  */
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <algorithm>
-#include <cassert>
-#include <iostream>
-#include <memory>
-#include <numeric>
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_properties.hpp>
-#include <raft/core/resources.hpp>
-#include <raft/neighbors/sample_filter_types.hpp>
-
-#include <vector>
-
 #include "bitonic.hpp"
 #include "compute_distance.hpp"
 #include "device_common.hpp"
@@ -37,10 +22,24 @@
 #include "search_plan.cuh"
 #include "topk_for_cagra/topk_core.cuh"  // TODO replace with raft topk if possible
 #include "utils.hpp"
+
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_properties.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/neighbors/sample_filter_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 #include <raft/util/cudart_utils.hpp>  // RAFT_CUDA_TRY_NOT_THROW is used TODO(tfeher): consider moving this to cuda_rt_essentials.hpp
 
+#include <algorithm>
+#include <cassert>
+#include <iostream>
+#include <memory>
+#include <numeric>
+#include <vector>
+
 namespace raft::neighbors::cagra::detail {
 namespace multi_cta_search {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_multi_kernel.cuh b/cpp/include/raft/neighbors/detail/cagra/search_multi_kernel.cuh
index f9bf525503..e4a30675bb 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_multi_kernel.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_multi_kernel.cuh
@@ -15,32 +15,33 @@
  */
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <algorithm>
-#include <cassert>
-#include <iostream>
-#include <memory>
-#include <numeric>
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resources.hpp>
-#include <raft/neighbors/sample_filter_types.hpp>
-#include <rmm/device_scalar.hpp>
-#include <rmm/device_uvector.hpp>
-#include <vector>
-
 #include "compute_distance.hpp"
 #include "device_common.hpp"
 #include "hashmap.hpp"
 #include "search_plan.cuh"
 #include "topk_for_cagra/topk_core.cuh"  //todo replace with raft kernel
 #include "utils.hpp"
+
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/matrix/select_k.cuh>
+#include <raft/neighbors/sample_filter_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 #include <raft/util/cudart_utils.hpp>  // RAFT_CUDA_TRY_NOT_THROW is used TODO(tfeher): consider moving this to cuda_rt_essentials.hpp
 
+#include <rmm/device_scalar.hpp>
+#include <rmm/device_uvector.hpp>
+
+#include <algorithm>
+#include <cassert>
+#include <iostream>
+#include <memory>
+#include <numeric>
+#include <vector>
+
 namespace raft::neighbors::cagra::detail {
 namespace multi_kernel_search {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_plan.cuh b/cpp/include/raft/neighbors/detail/cagra/search_plan.cuh
index 271a1f4955..11ef7e5211 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_plan.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_plan.cuh
@@ -17,7 +17,9 @@
 #pragma once
 
 #include "hashmap.hpp"
+
 #include <raft/core/resource/cuda_stream.hpp>
+#include <raft/neighbors/sample_filter_types.hpp>
 // #include "search_single_cta.cuh"
 // #include "topk_for_cagra/topk_core.cuh"
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_single_cta.cuh b/cpp/include/raft/neighbors/detail/cagra/search_single_cta.cuh
index 0b4fc2d47b..f1e74ee7a5 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_single_cta.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_single_cta.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,20 +15,6 @@
  */
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <algorithm>
-#include <cassert>
-#include <iostream>
-#include <memory>
-#include <numeric>
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_properties.hpp>
-#include <raft/core/resources.hpp>
-#include <rmm/device_uvector.hpp>
-#include <vector>
-
 #include "bitonic.hpp"
 #include "compute_distance.hpp"
 #include "device_common.hpp"
@@ -38,10 +24,25 @@
 #include "topk_by_radix.cuh"
 #include "topk_for_cagra/topk_core.cuh"  // TODO replace with raft topk
 #include "utils.hpp"
+
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_properties.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 #include <raft/util/cudart_utils.hpp>  // RAFT_CUDA_TRY_NOT_THROW is used TODO(tfeher): consider moving this to cuda_rt_essentials.hpp
 
+#include <rmm/device_uvector.hpp>
+
+#include <algorithm>
+#include <cassert>
+#include <iostream>
+#include <memory>
+#include <numeric>
+#include <vector>
+
 namespace raft::neighbors::cagra::detail {
 namespace single_cta_search {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/search_single_cta_kernel-inl.cuh b/cpp/include/raft/neighbors/detail/cagra/search_single_cta_kernel-inl.cuh
index 80b5b343b2..652115928b 100644
--- a/cpp/include/raft/neighbors/detail/cagra/search_single_cta_kernel-inl.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/search_single_cta_kernel-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,21 +15,6 @@
  */
 #pragma once
 
-#include <algorithm>
-#include <cassert>
-#include <cstdint>
-#include <iostream>
-#include <memory>
-#include <numeric>
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_properties.hpp>
-#include <raft/core/resources.hpp>
-#include <raft/neighbors/sample_filter_types.hpp>
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-#include <rmm/device_uvector.hpp>
-#include <vector>
-
 #include "bitonic.hpp"
 #include "compute_distance.hpp"
 #include "device_common.hpp"
@@ -38,10 +23,27 @@
 #include "topk_by_radix.cuh"
 #include "topk_for_cagra/topk_core.cuh"  // TODO replace with raft topk
 #include "utils.hpp"
+
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_properties.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/neighbors/sample_filter_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_rt_essentials.hpp>
 #include <raft/util/cudart_utils.hpp>  // RAFT_CUDA_TRY_NOT_THROW is used TODO(tfeher): consider moving this to cuda_rt_essentials.hpp
 
+#include <rmm/device_uvector.hpp>
+
+#include <algorithm>
+#include <cassert>
+#include <cstdint>
+#include <iostream>
+#include <memory>
+#include <numeric>
+#include <vector>
+
 namespace raft::neighbors::cagra::detail {
 namespace single_cta_search {
 
diff --git a/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk.h b/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk.h
index 92b9474047..ac1a746090 100644
--- a/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk.h
+++ b/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,6 +16,7 @@
 #pragma once
 
 #include <cuda_fp16.h>
+
 #include <stdint.h>
 
 namespace raft::neighbors::cagra::detail {
diff --git a/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk_core.cuh b/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk_core.cuh
index 8e04a2eb0d..aedeb1be67 100644
--- a/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk_core.cuh
+++ b/cpp/include/raft/neighbors/detail/cagra/topk_for_cagra/topk_core.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,10 @@
  */
 #pragma once
 #include "topk.h"
-#include <assert.h>
+
 #include <cub/cub.cuh>
+
+#include <assert.h>
 #include <float.h>
 #include <stdint.h>
 #include <stdio.h>
diff --git a/cpp/include/raft/neighbors/detail/cagra/utils.hpp b/cpp/include/raft/neighbors/detail/cagra/utils.hpp
index 5e57a9589f..7e403abe91 100644
--- a/cpp/include/raft/neighbors/detail/cagra/utils.hpp
+++ b/cpp/include/raft/neighbors/detail/cagra/utils.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,15 +15,18 @@
  */
 #pragma once
 
-#include <cfloat>
-#include <cstdint>
-#include <cuda.h>
-#include <cuda_fp16.h>
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/util/integer_utils.hpp>
+
 #include <rmm/mr/device/device_memory_resource.hpp>
+
+#include <cuda.h>
+#include <cuda_fp16.h>
+
+#include <cfloat>
+#include <cstdint>
 #include <type_traits>
 
 namespace raft::neighbors::cagra::detail {
diff --git a/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkBlock.cuh b/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkBlock.cuh
index 1f4308fa2f..0677b0596c 100644
--- a/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkBlock.cuh
+++ b/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkBlock.cuh
@@ -7,10 +7,11 @@
 
 #pragma once
 
-#include <cuda.h>
 #include <raft/neighbors/detail/faiss_select/MergeNetworkUtils.cuh>
 #include <raft/neighbors/detail/faiss_select/StaticUtils.h>
 
+#include <cuda.h>
+
 namespace raft::neighbors::detail::faiss_select {
 
 // Merge pairs of lists smaller than blockDim.x (NumThreads)
diff --git a/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkWarp.cuh b/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkWarp.cuh
index 3d87e284ed..64b1bada48 100644
--- a/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkWarp.cuh
+++ b/cpp/include/raft/neighbors/detail/faiss_select/MergeNetworkWarp.cuh
@@ -9,7 +9,6 @@
 
 #include <raft/neighbors/detail/faiss_select/MergeNetworkUtils.cuh>
 #include <raft/neighbors/detail/faiss_select/StaticUtils.h>
-
 #include <raft/util/cuda_utils.cuh>
 
 namespace raft::neighbors::detail::faiss_select {
diff --git a/cpp/include/raft/neighbors/detail/faiss_select/Select.cuh b/cpp/include/raft/neighbors/detail/faiss_select/Select.cuh
index fdb986ac69..445e638eaa 100644
--- a/cpp/include/raft/neighbors/detail/faiss_select/Select.cuh
+++ b/cpp/include/raft/neighbors/detail/faiss_select/Select.cuh
@@ -7,11 +7,10 @@
 
 #pragma once
 
+#include <raft/core/kvp.hpp>
 #include <raft/neighbors/detail/faiss_select/Comparators.cuh>
 #include <raft/neighbors/detail/faiss_select/MergeNetworkBlock.cuh>
 #include <raft/neighbors/detail/faiss_select/MergeNetworkWarp.cuh>
-
-#include <raft/core/kvp.hpp>
 #include <raft/util/cuda_utils.cuh>
 
 namespace raft::neighbors::detail::faiss_select {
diff --git a/cpp/include/raft/neighbors/detail/hnsw.hpp b/cpp/include/raft/neighbors/detail/hnsw.hpp
index 69478205a9..5deee3c2ba 100644
--- a/cpp/include/raft/neighbors/detail/hnsw.hpp
+++ b/cpp/include/raft/neighbors/detail/hnsw.hpp
@@ -18,13 +18,13 @@
 
 #include "hnsw_types.hpp"
 
-#include <cstdint>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resources.hpp>
 
+#include <hnswlib/hnswlib.h>
 #include <omp.h>
 
-#include <hnswlib/hnswlib.h>
+#include <cstdint>
 
 namespace raft::neighbors::hnsw::detail {
 
diff --git a/cpp/include/raft/neighbors/detail/hnsw_serialize.hpp b/cpp/include/raft/neighbors/detail/hnsw_serialize.hpp
index 8103ffc5ab..4c3728f8fc 100644
--- a/cpp/include/raft/neighbors/detail/hnsw_serialize.hpp
+++ b/cpp/include/raft/neighbors/detail/hnsw_serialize.hpp
@@ -19,8 +19,6 @@
 #include "../hnsw_types.hpp"
 #include "hnsw_types.hpp"
 
-#include <cstddef>
-#include <cstdint>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/mdarray.hpp>
 #include <raft/core/mdspan_types.hpp>
@@ -29,6 +27,8 @@
 #include <raft/core/serialize.hpp>
 #include <raft/neighbors/cagra_types.hpp>
 
+#include <cstddef>
+#include <cstdint>
 #include <fstream>
 #include <type_traits>
 
diff --git a/cpp/include/raft/neighbors/detail/hnsw_types.hpp b/cpp/include/raft/neighbors/detail/hnsw_types.hpp
index 94ade95965..9d35effd1a 100644
--- a/cpp/include/raft/neighbors/detail/hnsw_types.hpp
+++ b/cpp/include/raft/neighbors/detail/hnsw_types.hpp
@@ -17,13 +17,15 @@
 #pragma once
 
 #include "../hnsw_types.hpp"
-#include <memory>
+
 #include <raft/core/error.hpp>
 #include <raft/distance/distance_types.hpp>
 
-#include <cstdint>
 #include <hnswlib/hnswlib.h>
 #include <sys/types.h>
+
+#include <cstdint>
+#include <memory>
 #include <type_traits>
 
 namespace raft::neighbors::hnsw::detail {
diff --git a/cpp/include/raft/neighbors/detail/ivf_common.cuh b/cpp/include/raft/neighbors/detail/ivf_common.cuh
new file mode 100644
index 0000000000..ef7ae7c804
--- /dev/null
+++ b/cpp/include/raft/neighbors/detail/ivf_common.cuh
@@ -0,0 +1,326 @@
+/*
+ * Copyright (c) 2024, NVIDIA CORPORATION.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#pragma once
+
+#include <raft/linalg/unary_op.cuh>
+#include <raft/matrix/detail/select_warpsort.cuh>  // matrix::detail::select::warpsort::warp_sort_distributed
+
+#include <cub/cub.cuh>
+
+namespace raft::neighbors::ivf::detail {
+
+/**
+ * Default value returned by `search` when the `n_probes` is too small and top-k is too large.
+ * One may encounter it if the combined size of probed clusters is smaller than the requested
+ * number of results per query.
+ */
+template <typename IdxT>
+constexpr static IdxT kOutOfBoundsRecord = std::numeric_limits<IdxT>::max();
+
+template <typename T, typename IdxT, bool Ascending = true>
+struct dummy_block_sort_t {
+  using queue_t =
+    matrix::detail::select::warpsort::warp_sort_distributed<WarpSize, Ascending, T, IdxT>;
+  template <typename... Args>
+  __device__ dummy_block_sort_t(int k, Args...){};
+};
+
+/**
+ * For each query, we calculate a cumulative sum of the cluster sizes that we probe, and return that
+ * in chunk_indices. Essentially this is a segmented inclusive scan of the cluster sizes. The total
+ * number of samples per query (sum of the cluster sizes that we probe) is returned in n_samples.
+ */
+template <int BlockDim>
+__launch_bounds__(BlockDim) RAFT_KERNEL
+  calc_chunk_indices_kernel(uint32_t n_probes,
+                            const uint32_t* cluster_sizes,      // [n_clusters]
+                            const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
+                            uint32_t* chunk_indices,            // [n_queries, n_probes]
+                            uint32_t* n_samples                 // [n_queries]
+  )
+{
+  using block_scan = cub::BlockScan<uint32_t, BlockDim>;
+  __shared__ typename block_scan::TempStorage shm;
+
+  // locate the query data
+  clusters_to_probe += n_probes * blockIdx.x;
+  chunk_indices += n_probes * blockIdx.x;
+
+  // block scan
+  const uint32_t n_probes_aligned = Pow2<BlockDim>::roundUp(n_probes);
+  uint32_t total                  = 0;
+  for (uint32_t probe_ix = threadIdx.x; probe_ix < n_probes_aligned; probe_ix += BlockDim) {
+    auto label = probe_ix < n_probes ? clusters_to_probe[probe_ix] : 0u;
+    auto chunk = probe_ix < n_probes ? cluster_sizes[label] : 0u;
+    if (threadIdx.x == 0) { chunk += total; }
+    block_scan(shm).InclusiveSum(chunk, chunk, total);
+    __syncthreads();
+    if (probe_ix < n_probes) { chunk_indices[probe_ix] = chunk; }
+  }
+  // save the total size
+  if (threadIdx.x == 0) { n_samples[blockIdx.x] = total; }
+}
+
+struct calc_chunk_indices {
+ public:
+  struct configured {
+    void* kernel;
+    dim3 block_dim;
+    dim3 grid_dim;
+    uint32_t n_probes;
+
+    inline void operator()(const uint32_t* cluster_sizes,
+                           const uint32_t* clusters_to_probe,
+                           uint32_t* chunk_indices,
+                           uint32_t* n_samples,
+                           rmm::cuda_stream_view stream)
+    {
+      void* args[] =  // NOLINT
+        {&n_probes, &cluster_sizes, &clusters_to_probe, &chunk_indices, &n_samples};
+      RAFT_CUDA_TRY(cudaLaunchKernel(kernel, grid_dim, block_dim, args, 0, stream));
+    }
+  };
+
+  static inline auto configure(uint32_t n_probes, uint32_t n_queries) -> configured
+  {
+    return try_block_dim<1024>(n_probes, n_queries);
+  }
+
+ private:
+  template <int BlockDim>
+  static auto try_block_dim(uint32_t n_probes, uint32_t n_queries) -> configured
+  {
+    if constexpr (BlockDim >= WarpSize * 2) {
+      if (BlockDim >= n_probes * 2) { return try_block_dim<(BlockDim / 2)>(n_probes, n_queries); }
+    }
+    return {reinterpret_cast<void*>(calc_chunk_indices_kernel<BlockDim>),
+            dim3(BlockDim, 1, 1),
+            dim3(n_queries, 1, 1),
+            n_probes};
+  }
+};
+
+/**
+ * Look up the chunk id corresponding to the sample index.
+ *
+ * Each query vector was compared to all the vectors from n_probes clusters, and sample_ix is an
+ * ordered number of one of such vectors. This function looks up to which chunk it belongs,
+ * and returns the index within the chunk (which is also an index within a cluster).
+ *
+ * @param[inout] sample_ix
+ *   input: the offset of the sample in the batch;
+ *   output: the offset inside the chunk (probe) / selected cluster.
+ * @param[in] n_probes number of probes
+ * @param[in] chunk_indices offsets of the chunks within the batch [n_probes]
+ * @return chunk index (== n_probes when the input index is not in the valid range,
+ *    which can happen if there is not enough data to output in the selected clusters).
+ */
+__device__ inline auto find_chunk_ix(uint32_t& sample_ix,  // NOLINT
+                                     uint32_t n_probes,
+                                     const uint32_t* chunk_indices) -> uint32_t
+{
+  uint32_t ix_min = 0;
+  uint32_t ix_max = n_probes;
+  do {
+    uint32_t i = (ix_min + ix_max) / 2;
+    if (chunk_indices[i] <= sample_ix) {
+      ix_min = i + 1;
+    } else {
+      ix_max = i;
+    }
+  } while (ix_min < ix_max);
+  if (ix_min > 0) { sample_ix -= chunk_indices[ix_min - 1]; }
+  return ix_min;
+}
+
+template <int BlockDim, typename IdxT1, typename IdxT2 = uint32_t>
+__launch_bounds__(BlockDim) RAFT_KERNEL
+  postprocess_neighbors_kernel(IdxT1* neighbors_out,               // [n_queries, topk]
+                               const IdxT2* neighbors_in,          // [n_queries, topk]
+                               const IdxT1* const* db_indices,     // [n_clusters][..]
+                               const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
+                               const uint32_t* chunk_indices,      // [n_queries, n_probes]
+                               uint32_t n_queries,
+                               uint32_t n_probes,
+                               uint32_t topk)
+{
+  const uint64_t i        = threadIdx.x + BlockDim * uint64_t(blockIdx.x);
+  const uint32_t query_ix = i / uint64_t(topk);
+  if (query_ix >= n_queries) { return; }
+  const uint32_t k = i % uint64_t(topk);
+  neighbors_in += query_ix * topk;
+  neighbors_out += query_ix * topk;
+  chunk_indices += query_ix * n_probes;
+  clusters_to_probe += query_ix * n_probes;
+  uint32_t data_ix        = neighbors_in[k];
+  const uint32_t chunk_ix = find_chunk_ix(data_ix, n_probes, chunk_indices);
+  const bool valid        = chunk_ix < n_probes;
+  neighbors_out[k] =
+    valid ? db_indices[clusters_to_probe[chunk_ix]][data_ix] : kOutOfBoundsRecord<IdxT1>;
+}
+
+/**
+ * Transform found sample indices into the corresponding database indices
+ * (as stored in index.indices()).
+ * The sample indices are the record indices as they appear in the database view formed by the
+ * probed clusters / defined by the `chunk_indices`.
+ * We assume the searched sample sizes (for a single query) fit into `uint32_t`.
+ */
+template <typename IdxT1, typename IdxT2 = uint32_t>
+void postprocess_neighbors(IdxT1* neighbors_out,               // [n_queries, topk]
+                           const IdxT2* neighbors_in,          // [n_queries, topk]
+                           const IdxT1* const* db_indices,     // [n_clusters][..]
+                           const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
+                           const uint32_t* chunk_indices,      // [n_queries, n_probes]
+                           uint32_t n_queries,
+                           uint32_t n_probes,
+                           uint32_t topk,
+                           rmm::cuda_stream_view stream)
+{
+  constexpr int kPNThreads = 256;
+  const int pn_blocks      = raft::div_rounding_up_unsafe<size_t>(n_queries * topk, kPNThreads);
+  postprocess_neighbors_kernel<kPNThreads, IdxT1, IdxT2>
+    <<<pn_blocks, kPNThreads, 0, stream>>>(neighbors_out,
+                                           neighbors_in,
+                                           db_indices,
+                                           clusters_to_probe,
+                                           chunk_indices,
+                                           n_queries,
+                                           n_probes,
+                                           topk);
+}
+
+/**
+ * Post-process the scores depending on the metric type;
+ * translate the element type if necessary.
+ */
+template <typename ScoreInT, typename ScoreOutT = float>
+void postprocess_distances(ScoreOutT* out,      // [n_queries, topk]
+                           const ScoreInT* in,  // [n_queries, topk]
+                           distance::DistanceType metric,
+                           uint32_t n_queries,
+                           uint32_t topk,
+                           float scaling_factor,
+                           bool account_for_max_close,
+                           rmm::cuda_stream_view stream)
+{
+  constexpr bool needs_cast = !std::is_same<ScoreInT, ScoreOutT>::value;
+  const bool needs_copy     = ((void*)in) != ((void*)out);
+  size_t len                = size_t(n_queries) * size_t(topk);
+  switch (metric) {
+    case distance::DistanceType::L2Unexpanded:
+    case distance::DistanceType::L2Expanded: {
+      if (scaling_factor != 1.0) {
+        linalg::unaryOp(
+          out,
+          in,
+          len,
+          raft::compose_op(raft::mul_const_op<ScoreOutT>{scaling_factor * scaling_factor},
+                           raft::cast_op<ScoreOutT>{}),
+          stream);
+      } else if (needs_cast || needs_copy) {
+        linalg::unaryOp(out, in, len, raft::cast_op<ScoreOutT>{}, stream);
+      }
+    } break;
+    case distance::DistanceType::L2SqrtUnexpanded:
+    case distance::DistanceType::L2SqrtExpanded: {
+      if (scaling_factor != 1.0) {
+        linalg::unaryOp(out,
+                        in,
+                        len,
+                        raft::compose_op{raft::mul_const_op<ScoreOutT>{scaling_factor},
+                                         raft::sqrt_op{},
+                                         raft::cast_op<ScoreOutT>{}},
+                        stream);
+      } else if (needs_cast) {
+        linalg::unaryOp(
+          out, in, len, raft::compose_op{raft::sqrt_op{}, raft::cast_op<ScoreOutT>{}}, stream);
+      } else {
+        linalg::unaryOp(out, in, len, raft::sqrt_op{}, stream);
+      }
+    } break;
+    case distance::DistanceType::InnerProduct: {
+      float factor = (account_for_max_close ? -1.0 : 1.0) * scaling_factor * scaling_factor;
+      if (factor != 1.0) {
+        linalg::unaryOp(
+          out,
+          in,
+          len,
+          raft::compose_op(raft::mul_const_op<ScoreOutT>{factor}, raft::cast_op<ScoreOutT>{}),
+          stream);
+      } else if (needs_cast || needs_copy) {
+        linalg::unaryOp(out, in, len, raft::cast_op<ScoreOutT>{}, stream);
+      }
+    } break;
+    default: RAFT_FAIL("Unexpected metric.");
+  }
+}
+
+/** Update the state of the dependent index members. */
+template <typename Index>
+void recompute_internal_state(const raft::resources& res, Index& index)
+{
+  auto stream  = resource::get_cuda_stream(res);
+  auto tmp_res = resource::get_workspace_resource(res);
+  rmm::device_uvector<uint32_t> sorted_sizes(index.n_lists(), stream, tmp_res);
+
+  // Actualize the list pointers
+  auto data_ptrs = index.data_ptrs();
+  auto inds_ptrs = index.inds_ptrs();
+  for (uint32_t label = 0; label < index.n_lists(); label++) {
+    auto& list          = index.lists()[label];
+    const auto data_ptr = list ? list->data.data_handle() : nullptr;
+    const auto inds_ptr = list ? list->indices.data_handle() : nullptr;
+    copy(&data_ptrs(label), &data_ptr, 1, stream);
+    copy(&inds_ptrs(label), &inds_ptr, 1, stream);
+  }
+
+  // Sort the cluster sizes in the descending order.
+  int begin_bit             = 0;
+  int end_bit               = sizeof(uint32_t) * 8;
+  size_t cub_workspace_size = 0;
+  cub::DeviceRadixSort::SortKeysDescending(nullptr,
+                                           cub_workspace_size,
+                                           index.list_sizes().data_handle(),
+                                           sorted_sizes.data(),
+                                           index.n_lists(),
+                                           begin_bit,
+                                           end_bit,
+                                           stream);
+  rmm::device_buffer cub_workspace(cub_workspace_size, stream, tmp_res);
+  cub::DeviceRadixSort::SortKeysDescending(cub_workspace.data(),
+                                           cub_workspace_size,
+                                           index.list_sizes().data_handle(),
+                                           sorted_sizes.data(),
+                                           index.n_lists(),
+                                           begin_bit,
+                                           end_bit,
+                                           stream);
+  // copy the results to CPU
+  std::vector<uint32_t> sorted_sizes_host(index.n_lists());
+  copy(sorted_sizes_host.data(), sorted_sizes.data(), index.n_lists(), stream);
+  resource::sync_stream(res);
+
+  // accumulate the sorted cluster sizes
+  auto accum_sorted_sizes = index.accum_sorted_sizes();
+  accum_sorted_sizes(0)   = 0;
+  for (uint32_t label = 0; label < sorted_sizes_host.size(); label++) {
+    accum_sorted_sizes(label + 1) = accum_sorted_sizes(label) + sorted_sizes_host[label];
+  }
+}
+
+}  // namespace raft::neighbors::ivf::detail
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_build.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_build.cuh
index fa11d9236f..55184cc615 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_build.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_build.cuh
@@ -26,6 +26,7 @@
 #include <raft/linalg/add.cuh>
 #include <raft/linalg/map.cuh>
 #include <raft/linalg/norm.cuh>
+#include <raft/neighbors/detail/ivf_common.cuh>
 #include <raft/neighbors/ivf_flat_codepacker.hpp>
 #include <raft/neighbors/ivf_flat_types.hpp>
 #include <raft/neighbors/ivf_list.hpp>
@@ -75,7 +76,7 @@ auto clone(const raft::resources& res, const index<T, IdxT>& source) -> index<T,
   target.lists() = source.lists();
 
   // Make sure the device pointers point to the new lists
-  target.recompute_internal_state(res);
+  ivf::detail::recompute_internal_state(res, target);
 
   return target;
 }
@@ -262,7 +263,7 @@ void extend(raft::resources const& handle,
     }
   }
   // Update the pointers and the sizes
-  index->recompute_internal_state(handle);
+  ivf::detail::recompute_internal_state(handle, *index);
   // Copy the old sizes, so we can start from the current state of the index;
   // we'll rebuild the `list_sizes_ptr` in the following kernel, using it as an atomic counter.
   raft::copy(list_sizes_ptr, old_list_sizes_dev.data_handle(), n_lists, stream);
@@ -355,6 +356,8 @@ inline auto build(raft::resources const& handle,
   RAFT_EXPECTS(n_rows >= params.n_lists, "number of rows can't be less than n_lists");
 
   index<T, IdxT> index(handle, params, dim);
+  utils::memzero(
+    index.accum_sorted_sizes().data_handle(), index.accum_sorted_sizes().size(), stream);
   utils::memzero(index.list_sizes().data_handle(), index.list_sizes().size(), stream);
   utils::memzero(index.data_ptrs().data_handle(), index.data_ptrs().size(), stream);
   utils::memzero(index.inds_ptrs().data_handle(), index.inds_ptrs().size(), stream);
@@ -442,7 +445,7 @@ inline void fill_refinement_index(raft::resources const& handle,
     ivf::resize_list(handle, lists[label], list_device_spec, n_candidates, uint32_t(0));
   }
   // Update the pointers and the sizes
-  refinement_index->recompute_internal_state(handle);
+  ivf::detail::recompute_internal_state(handle, *refinement_index);
 
   RAFT_CUDA_TRY(cudaMemsetAsync(list_sizes_ptr, 0, n_lists * sizeof(uint32_t), stream));
 
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-ext.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-ext.cuh
index 58e94ee7aa..7c2d1d2157 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-ext.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-ext.cuh
@@ -16,18 +16,22 @@
 
 #pragma once
 
-#include <cstdint>                                 // uintX_t
 #include <raft/neighbors/ivf_flat_types.hpp>       // raft::neighbors::ivf_flat::index
 #include <raft/neighbors/sample_filter_types.hpp>  // none_ivf_sample_filter
 #include <raft/util/raft_explicit.hpp>             // RAFT_EXPLICIT
-#include <rmm/cuda_stream_view.hpp>                // rmm:cuda_stream_view
+
+#include <rmm/cuda_stream_view.hpp>  // rmm:cuda_stream_view
 
 #include <cuda_fp16.h>
 
+#include <cstdint>  // uintX_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::ivf_flat::detail {
 
+auto RAFT_WEAK_FUNCTION is_local_topk_feasible(uint32_t k) -> bool;
+
 template <typename T, typename AccT, typename IdxT, typename IvfSampleFilterT>
 void ivfflat_interleaved_scan(const raft::neighbors::ivf_flat::index<T, IdxT>& index,
                               const T* queries,
@@ -37,6 +41,8 @@ void ivfflat_interleaved_scan(const raft::neighbors::ivf_flat::index<T, IdxT>& i
                               const raft::distance::DistanceType metric,
                               const uint32_t n_probes,
                               const uint32_t k,
+                              const uint32_t max_samples,
+                              const uint32_t* chunk_indices,
                               const bool select_min,
                               IvfSampleFilterT sample_filter,
                               IdxT* neighbors,
@@ -60,6 +66,8 @@ void ivfflat_interleaved_scan(const raft::neighbors::ivf_flat::index<T, IdxT>& i
     const raft::distance::DistanceType metric,                                                  \
     const uint32_t n_probes,                                                                    \
     const uint32_t k,                                                                           \
+    const uint32_t max_samples,                                                                 \
+    const uint32_t* chunk_indices,                                                              \
     const bool select_min,                                                                      \
     IvfSampleFilterT sample_filter,                                                             \
     IdxT* neighbors,                                                                            \
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-inl.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-inl.cuh
index 51cd2876d8..6fc528e26b 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-inl.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_interleaved_scan-inl.cuh
@@ -20,6 +20,7 @@
 #include <raft/core/operators.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/matrix/detail/select_warpsort.cuh>
+#include <raft/neighbors/detail/ivf_common.cuh>
 #include <raft/neighbors/ivf_flat_types.hpp>
 #include <raft/neighbors/sample_filter_types.hpp>
 #include <raft/spatial/knn/detail/ann_utils.cuh>
@@ -28,6 +29,7 @@
 #include <raft/util/integer_utils.hpp>
 #include <raft/util/pow2_utils.cuh>
 #include <raft/util/vectorized.cuh>
+
 #include <rmm/cuda_stream_view.hpp>
 
 namespace raft::neighbors::ivf_flat::detail {
@@ -36,6 +38,11 @@ using namespace raft::spatial::knn::detail;  // NOLINT
 
 constexpr int kThreadsPerBlock = 128;
 
+auto RAFT_WEAK_FUNCTION is_local_topk_feasible(uint32_t k) -> bool
+{
+  return k <= matrix::detail::select::warpsort::kMaxCapacity;
+}
+
 /**
  * @brief Copy `n` elements per block from one place to another.
  *
@@ -627,6 +634,23 @@ struct loadAndComputeDist<kUnroll, Lambda, 1, int8_t, int32_t> {
   }
 };
 
+// switch to dummy blocksort when Capacity is 0 this explicit dummy is chosen
+// to support access to warpsort constants like ::queue_t::kDummy
+template <int Capacity, bool Ascending, typename T, typename IdxT>
+struct flat_block_sort {
+  using type = matrix::detail::select::warpsort::
+    block_sort<matrix::detail::select::warpsort::warp_sort_filtered, Capacity, Ascending, T, IdxT>;
+};
+
+template <typename T, bool Ascending, typename IdxT>
+struct flat_block_sort<0, Ascending, T, IdxT>
+  : ivf::detail::dummy_block_sort_t<T, IdxT, Ascending> {
+  using type = ivf::detail::dummy_block_sort_t<T, IdxT, Ascending>;
+};
+
+template <int Capacity, bool Ascending, typename T, typename IdxT>
+using block_sort_t = typename flat_block_sort<Capacity, Ascending, T, IdxT>::type;
+
 /**
  * Scan clusters for nearest neighbors of the query vectors.
  * See `ivfflat_interleaved_scan` for more information.
@@ -672,12 +696,15 @@ RAFT_KERNEL __launch_bounds__(kThreadsPerBlock)
                           const uint32_t queries_offset,
                           const uint32_t n_probes,
                           const uint32_t k,
+                          const uint32_t max_samples,
+                          const uint32_t* chunk_indices,
                           const uint32_t dim,
                           IvfSampleFilterT sample_filter,
                           IdxT* neighbors,
                           float* distances)
 {
   extern __shared__ __align__(256) uint8_t interleaved_scan_kernel_smem[];
+  constexpr bool kManageLocalTopK = Capacity > 0;
   // Using shared memory for the (part of the) query;
   // This allows to save on global memory bandwidth when reading index and query
   // data at the same time.
@@ -687,8 +714,13 @@ RAFT_KERNEL __launch_bounds__(kThreadsPerBlock)
   {
     const int query_id = blockIdx.y;
     query += query_id * dim;
-    neighbors += query_id * k * gridDim.x + blockIdx.x * k;
-    distances += query_id * k * gridDim.x + blockIdx.x * k;
+    if constexpr (kManageLocalTopK) {
+      neighbors += query_id * k * gridDim.x + blockIdx.x * k;
+      distances += query_id * k * gridDim.x + blockIdx.x * k;
+    } else {
+      distances += query_id * uint64_t(max_samples);
+      chunk_indices += (n_probes * query_id);
+    }
     coarse_index += query_id * n_probes;
   }
 
@@ -696,14 +728,8 @@ RAFT_KERNEL __launch_bounds__(kThreadsPerBlock)
   copy_vectorized(query_shared, query, std::min(dim, query_smem_elems));
   __syncthreads();
 
-  using block_sort_t = matrix::detail::select::warpsort::block_sort<
-    matrix::detail::select::warpsort::warp_sort_filtered,
-    Capacity,
-    Ascending,
-    float,
-    IdxT>;
-  block_sort_t queue(k);
-
+  using local_topk_t = block_sort_t<Capacity, Ascending, float, IdxT>;
+  local_topk_t queue(k);
   {
     using align_warp  = Pow2<WarpSize>;
     const int lane_id = align_warp::mod(threadIdx.x);
@@ -725,6 +751,13 @@ RAFT_KERNEL __launch_bounds__(kThreadsPerBlock)
       const uint32_t num_groups =
         align_warp::div(list_length + align_warp::Mask);  // ceildiv by power of 2
 
+      uint32_t sample_offset = 0;
+      if constexpr (!kManageLocalTopK) {
+        if (probe_id > 0) { sample_offset = chunk_indices[probe_id - 1]; }
+        assert(list_length == chunk_indices[probe_id] - sample_offset);
+        assert(sample_offset + list_length <= max_samples);
+      }
+
       constexpr int kUnroll        = WarpSize / Veclen;
       constexpr uint32_t kNumWarps = kThreadsPerBlock / WarpSize;
       // Every warp reads WarpSize vectors and computes the distances to them.
@@ -771,17 +804,33 @@ RAFT_KERNEL __launch_bounds__(kThreadsPerBlock)
         }
 
         // Enqueue one element per thread
-        const float val  = valid ? static_cast<float>(dist) : block_sort_t::queue_t::kDummy;
-        const size_t idx = valid ? static_cast<size_t>(list_indices_ptrs[list_id][vec_id]) : 0;
-        queue.add(val, idx);
+        const float val = valid ? static_cast<float>(dist) : local_topk_t::queue_t::kDummy;
+        if constexpr (kManageLocalTopK) {
+          const size_t idx = valid ? static_cast<size_t>(list_indices_ptrs[list_id][vec_id]) : 0;
+          queue.add(val, idx);
+        } else {
+          if (vec_id < list_length) distances[sample_offset + vec_id] = val;
+        }
+      }
+
+      // fill up unused slots for current query
+      if constexpr (!kManageLocalTopK) {
+        if (probe_id + 1 == n_probes) {
+          for (uint32_t i = threadIdx.x + sample_offset + list_length; i < max_samples;
+               i += blockDim.x) {
+            distances[i] = local_topk_t::queue_t::kDummy;
+          }
+        }
       }
     }
   }
 
   // finalize and store selected neighbours
-  __syncthreads();
-  queue.done(interleaved_scan_kernel_smem);
-  queue.store(distances, neighbors, post_process);
+  if constexpr (kManageLocalTopK) {
+    __syncthreads();
+    queue.done(interleaved_scan_kernel_smem);
+    queue.store(distances, neighbors, post_process);
+  }
 }
 
 /**
@@ -821,6 +870,8 @@ void launch_kernel(Lambda lambda,
                    const uint32_t queries_offset,
                    const uint32_t n_probes,
                    const uint32_t k,
+                   const uint32_t max_samples,
+                   const uint32_t* chunk_indices,
                    IvfSampleFilterT sample_filter,
                    IdxT* neighbors,
                    float* distances,
@@ -841,12 +892,15 @@ void launch_kernel(Lambda lambda,
   const int max_query_smem = 16384;
   int query_smem_elems =
     std::min<int>(max_query_smem / sizeof(T), Pow2<Veclen * WarpSize>::roundUp(index.dim()));
-  int smem_size              = query_smem_elems * sizeof(T);
-  constexpr int kSubwarpSize = std::min<int>(Capacity, WarpSize);
-  auto block_merge_mem =
-    raft::matrix::detail::select::warpsort::calc_smem_size_for_block_wide<float, IdxT>(
-      kThreadsPerBlock / kSubwarpSize, k);
-  smem_size += std::max<int>(smem_size, block_merge_mem);
+  int smem_size = query_smem_elems * sizeof(T);
+
+  if constexpr (Capacity > 0) {
+    constexpr int kSubwarpSize = std::min<int>(Capacity, WarpSize);
+    auto block_merge_mem =
+      raft::matrix::detail::select::warpsort::calc_smem_size_for_block_wide<float, IdxT>(
+        kThreadsPerBlock / kSubwarpSize, k);
+    smem_size += std::max<int>(smem_size, block_merge_mem);
+  }
 
   // power-of-two less than cuda limit (for better addr alignment)
   constexpr uint32_t kMaxGridY = 32768;
@@ -879,13 +933,20 @@ void launch_kernel(Lambda lambda,
                                                         queries_offset + query_offset,
                                                         n_probes,
                                                         k,
+                                                        max_samples,
+                                                        chunk_indices,
                                                         index.dim(),
                                                         sample_filter,
                                                         neighbors,
                                                         distances);
     queries += grid_dim_y * index.dim();
-    neighbors += grid_dim_y * grid_dim_x * k;
-    distances += grid_dim_y * grid_dim_x * k;
+    if constexpr (Capacity > 0) {
+      neighbors += grid_dim_y * grid_dim_x * k;
+      distances += grid_dim_y * grid_dim_x * k;
+    } else {
+      distances += grid_dim_y * max_samples;
+      chunk_indices += grid_dim_y * n_probes;
+    }
     coarse_index += grid_dim_y * n_probes;
   }
 }
@@ -1010,16 +1071,22 @@ struct select_interleaved_scan_kernel {
    * two parameters and ends with both values equal to 1.
    */
   template <typename... Args>
-  static inline void run(int capacity, int veclen, bool select_min, Args&&... args)
+  static inline void run(int k_max, int veclen, bool select_min, Args&&... args)
   {
+    if constexpr (Capacity > 0) {
+      if (k_max == 0 || k_max > Capacity) {
+        return select_interleaved_scan_kernel<T, AccT, IdxT, IvfSampleFilterT, 0, Veclen>::run(
+          k_max, veclen, select_min, std::forward<Args>(args)...);
+      }
+    }
     if constexpr (Capacity > 1) {
-      if (capacity * 2 <= Capacity) {
+      if (k_max * 2 <= Capacity) {
         return select_interleaved_scan_kernel<T,
                                               AccT,
                                               IdxT,
                                               IvfSampleFilterT,
                                               Capacity / 2,
-                                              Veclen>::run(capacity,
+                                              Veclen>::run(k_max,
                                                            veclen,
                                                            select_min,
                                                            std::forward<Args>(args)...);
@@ -1028,14 +1095,14 @@ struct select_interleaved_scan_kernel {
     if constexpr (Veclen > 1) {
       if (veclen % Veclen != 0) {
         return select_interleaved_scan_kernel<T, AccT, IdxT, IvfSampleFilterT, Capacity, 1>::run(
-          capacity, 1, select_min, std::forward<Args>(args)...);
+          k_max, 1, select_min, std::forward<Args>(args)...);
       }
     }
     // NB: this is the limitation of the warpsort structures that use a huge number of
     //     registers (used in the main kernel here).
-    RAFT_EXPECTS(capacity == Capacity,
-                 "Capacity must be power-of-two not bigger than the maximum allowed size "
-                 "matrix::detail::select::warpsort::kMaxCapacity (%d).",
+    RAFT_EXPECTS(Capacity == 0 || k_max == Capacity,
+                 "Capacity must be either 0 or a power-of-two not bigger than the maximum "
+                 "allowed size matrix::detail::select::warpsort::kMaxCapacity (%d).",
                  matrix::detail::select::warpsort::kMaxCapacity);
     RAFT_EXPECTS(
       veclen == Veclen,
@@ -1090,6 +1157,8 @@ void ivfflat_interleaved_scan(const index<T, IdxT>& index,
                               const raft::distance::DistanceType metric,
                               const uint32_t n_probes,
                               const uint32_t k,
+                              const uint32_t max_samples,
+                              const uint32_t* chunk_indices,
                               const bool select_min,
                               IvfSampleFilterT sample_filter,
                               IdxT* neighbors,
@@ -1112,6 +1181,8 @@ void ivfflat_interleaved_scan(const index<T, IdxT>& index,
                                                                                queries_offset,
                                                                                n_probes,
                                                                                k,
+                                                                               max_samples,
+                                                                               chunk_indices,
                                                                                filter_adapter,
                                                                                neighbors,
                                                                                distances,
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_search-ext.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_search-ext.cuh
index d7dfe5c363..350b82ede7 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_search-ext.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_search-ext.cuh
@@ -16,13 +16,14 @@
 
 #pragma once
 
-#include <cstdint>                                 // uintX_t
 #include <raft/neighbors/ivf_flat_types.hpp>       // raft::neighbors::ivf_flat::index
 #include <raft/neighbors/sample_filter_types.hpp>  // none_ivf_sample_filter
 #include <raft/util/raft_explicit.hpp>             // RAFT_EXPLICIT
 
 #include <cuda_fp16.h>
 
+#include <cstdint>  // uintX_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::ivf_flat::detail {
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_search-inl.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_search-inl.cuh
index 29d521566d..98bdeda42f 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_search-inl.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_search-inl.cuh
@@ -24,11 +24,13 @@
 #include <raft/linalg/norm.cuh>                                 // raft::linalg::norm
 #include <raft/linalg/unary_op.cuh>                             // raft::linalg::unary_op
 #include <raft/matrix/detail/select_k.cuh>                      // matrix::detail::select_k
+#include <raft/neighbors/detail/ivf_common.cuh>                 // raft::neighbors::detail::ivf
 #include <raft/neighbors/detail/ivf_flat_interleaved_scan.cuh>  // interleaved_scan
 #include <raft/neighbors/ivf_flat_types.hpp>                    // raft::neighbors::ivf_flat::index
 #include <raft/neighbors/sample_filter_types.hpp>               // none_ivf_sample_filter
 #include <raft/spatial/knn/detail/ann_utils.cuh>                // utils::mapping
-#include <rmm/mr/device/per_device_resource.hpp>                // rmm::device_memory_resource
+
+#include <rmm/mr/device/per_device_resource.hpp>  // rmm::device_memory_resource
 
 namespace raft::neighbors::ivf_flat::detail {
 
@@ -42,6 +44,7 @@ void search_impl(raft::resources const& handle,
                  uint32_t queries_offset,
                  uint32_t k,
                  uint32_t n_probes,
+                 uint32_t max_samples,
                  bool select_min,
                  IdxT* neighbors,
                  AccT* distances,
@@ -49,18 +52,27 @@ void search_impl(raft::resources const& handle,
                  IvfSampleFilterT sample_filter)
 {
   auto stream = resource::get_cuda_stream(handle);
+
+  std::size_t n_queries_probes = std::size_t(n_queries) * std::size_t(n_probes);
+
   // The norm of query
   rmm::device_uvector<float> query_norm_dev(n_queries, stream, search_mr);
   // The distance value of cluster(list) and queries
   rmm::device_uvector<float> distance_buffer_dev(n_queries * index.n_lists(), stream, search_mr);
   // The topk distance value of cluster(list) and queries
-  rmm::device_uvector<float> coarse_distances_dev(n_queries * n_probes, stream, search_mr);
+  rmm::device_uvector<float> coarse_distances_dev(n_queries_probes, stream, search_mr);
   // The topk  index of cluster(list) and queries
-  rmm::device_uvector<uint32_t> coarse_indices_dev(n_queries * n_probes, stream, search_mr);
+  rmm::device_uvector<uint32_t> coarse_indices_dev(n_queries_probes, stream, search_mr);
+
+  // Optional structures if postprocessing is required
   // The topk distance value of candidate vectors from each cluster(list)
-  rmm::device_uvector<AccT> refined_distances_dev(n_queries * n_probes * k, stream, search_mr);
+  rmm::device_uvector<AccT> distances_tmp_dev(0, stream, search_mr);
   // The topk index of candidate vectors from each cluster(list)
-  rmm::device_uvector<IdxT> refined_indices_dev(n_queries * n_probes * k, stream, search_mr);
+  rmm::device_uvector<IdxT> indices_tmp_dev(0, stream, search_mr);
+  // Number of samples for each query
+  rmm::device_uvector<uint32_t> num_samples(0, stream, search_mr);
+  // Offsets per probe for each query
+  rmm::device_uvector<uint32_t> chunk_index(0, stream, search_mr);
 
   size_t float_query_size;
   if constexpr (std::is_integral_v<T>) {
@@ -139,9 +151,6 @@ void search_impl(raft::resources const& handle,
   RAFT_LOG_TRACE_VEC(coarse_indices_dev.data(), n_probes);
   RAFT_LOG_TRACE_VEC(coarse_distances_dev.data(), n_probes);
 
-  auto distances_dev_ptr = refined_distances_dev.data();
-  auto indices_dev_ptr   = refined_indices_dev.data();
-
   uint32_t grid_dim_x = 0;
   if (n_probes > 1) {
     // query the gridDimX size to store probes topK output
@@ -154,6 +163,8 @@ void search_impl(raft::resources const& handle,
       index.metric(),
       n_probes,
       k,
+      0,
+      nullptr,
       select_min,
       sample_filter,
       nullptr,
@@ -164,9 +175,30 @@ void search_impl(raft::resources const& handle,
     grid_dim_x = 1;
   }
 
-  if (grid_dim_x == 1) {
-    distances_dev_ptr = distances;
-    indices_dev_ptr   = neighbors;
+  auto distances_dev_ptr = distances;
+  auto indices_dev_ptr   = neighbors;
+
+  bool manage_local_topk = is_local_topk_feasible(k);
+  if (!manage_local_topk || grid_dim_x > 1) {
+    if (!manage_local_topk) {
+      num_samples.resize(n_queries, stream);
+      chunk_index.resize(n_queries_probes, stream);
+
+      ivf::detail::calc_chunk_indices::configure(n_probes, n_queries)(
+        index.list_sizes().data_handle(),
+        coarse_indices_dev.data(),
+        chunk_index.data(),
+        num_samples.data(),
+        stream);
+    }
+
+    auto target_size = std::size_t(n_queries) * (manage_local_topk ? grid_dim_x * k : max_samples);
+
+    distances_tmp_dev.resize(target_size, stream);
+    if (manage_local_topk) indices_tmp_dev.resize(target_size, stream);
+
+    distances_dev_ptr = distances_tmp_dev.data();
+    indices_dev_ptr   = indices_tmp_dev.data();
   }
 
   ivfflat_interleaved_scan<T, typename utils::config<T>::value_t, IdxT, IvfSampleFilterT>(
@@ -178,6 +210,8 @@ void search_impl(raft::resources const& handle,
     index.metric(),
     n_probes,
     k,
+    max_samples,
+    chunk_index.data(),
     select_min,
     sample_filter,
     indices_dev_ptr,
@@ -186,19 +220,34 @@ void search_impl(raft::resources const& handle,
     stream);
 
   RAFT_LOG_TRACE_VEC(distances_dev_ptr, 2 * k);
-  RAFT_LOG_TRACE_VEC(indices_dev_ptr, 2 * k);
+  if (indices_dev_ptr != nullptr) { RAFT_LOG_TRACE_VEC(indices_dev_ptr, 2 * k); }
 
   // Merge topk values from different blocks
-  if (grid_dim_x > 1) {
+  if (!manage_local_topk || grid_dim_x > 1) {
     matrix::detail::select_k<AccT, IdxT>(handle,
-                                         refined_distances_dev.data(),
-                                         refined_indices_dev.data(),
+                                         distances_tmp_dev.data(),
+                                         indices_tmp_dev.data(),
                                          n_queries,
-                                         k * grid_dim_x,
+                                         manage_local_topk ? (k * grid_dim_x) : max_samples,
                                          k,
                                          distances,
                                          neighbors,
                                          select_min);
+
+    if (!manage_local_topk) {
+      // post process distances && neighbor IDs
+      ivf::detail::postprocess_distances(
+        distances, distances, index.metric(), n_queries, k, 1.0, false, stream);
+      ivf::detail::postprocess_neighbors(neighbors,
+                                         neighbors,
+                                         index.inds_ptrs().data_handle(),
+                                         coarse_indices_dev.data(),
+                                         chunk_index.data(),
+                                         n_queries,
+                                         n_probes,
+                                         k,
+                                         stream);
+    }
   }
 }
 
@@ -223,14 +272,28 @@ inline void search(raft::resources const& handle,
   if (mr == nullptr) { mr = rmm::mr::get_current_device_resource(); }
   RAFT_EXPECTS(params.n_probes > 0,
                "n_probes (number of clusters to probe in the search) must be positive.");
-  auto n_probes = std::min<uint32_t>(params.n_probes, index.n_lists());
+  auto n_probes          = std::min<uint32_t>(params.n_probes, index.n_lists());
+  bool manage_local_topk = is_local_topk_feasible(k);
+
+  uint32_t max_samples = 0;
+  if (!manage_local_topk) {
+    IdxT ms =
+      Pow2<128 / sizeof(float)>::roundUp(std::max<IdxT>(index.accum_sorted_sizes()(n_probes), k));
+    RAFT_EXPECTS(ms <= IdxT(std::numeric_limits<uint32_t>::max()),
+                 "The maximum sample size is too big.");
+    max_samples = ms;
+  }
 
   // a batch size heuristic: try to keep the workspace within the specified size
-  constexpr uint32_t kExpectedWsSize = 1024 * 1024 * 1024;
+  constexpr uint64_t kExpectedWsSize = 1024 * 1024 * 1024;
+  uint64_t max_ws_size = std::min(resource::get_workspace_free_bytes(handle), kExpectedWsSize);
+
+  uint64_t ws_size_per_query = 4ull * (2 * n_probes + index.n_lists() + index.dim() + 1) +
+                               (manage_local_topk ? ((sizeof(IdxT) + 4) * n_probes * k)
+                                                  : (4ull * (max_samples + n_probes + 1)));
+
   const uint32_t max_queries =
-    std::min<uint32_t>(n_queries,
-                       raft::div_rounding_up_safe<uint64_t>(
-                         kExpectedWsSize, 16ull * uint64_t{n_probes} * k + 4ull * index.dim()));
+    std::min<uint32_t>(n_queries, raft::div_rounding_up_safe(max_ws_size, ws_size_per_query));
 
   for (uint32_t offset_q = 0; offset_q < n_queries; offset_q += max_queries) {
     uint32_t queries_batch = min(max_queries, n_queries - offset_q);
@@ -242,6 +305,7 @@ inline void search(raft::resources const& handle,
                                                   offset_q,
                                                   k,
                                                   n_probes,
+                                                  max_samples,
                                                   raft::distance::is_min_close(index.metric()),
                                                   neighbors + offset_q * k,
                                                   distances + offset_q * k,
diff --git a/cpp/include/raft/neighbors/detail/ivf_flat_serialize.cuh b/cpp/include/raft/neighbors/detail/ivf_flat_serialize.cuh
index aaf48ae830..3897b83aa6 100644
--- a/cpp/include/raft/neighbors/detail/ivf_flat_serialize.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_flat_serialize.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,6 +20,7 @@
 #include <raft/core/mdarray.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/serialize.hpp>
+#include <raft/neighbors/detail/ivf_flat_build.cuh>
 #include <raft/neighbors/ivf_flat_types.hpp>
 #include <raft/neighbors/ivf_list.hpp>
 #include <raft/neighbors/ivf_list_types.hpp>
@@ -153,7 +154,7 @@ auto deserialize(raft::resources const& handle, std::istream& is) -> index<T, Id
   }
   resource::sync_stream(handle);
 
-  index_.recompute_internal_state(handle);
+  ivf::detail::recompute_internal_state(handle, index_);
 
   return index_;
 }
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_build.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_build.cuh
index 2dfb261f32..b7796d52fa 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_build.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_build.cuh
@@ -16,18 +16,13 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <raft/neighbors/detail/ivf_pq_codepacking.cuh>
-#include <raft/neighbors/ivf_list.hpp>
-#include <raft/neighbors/ivf_pq_types.hpp>
-
 #include <raft/cluster/kmeans_balanced.cuh>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/logger.hpp>
 #include <raft/core/nvtx.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/device_memory_resource.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/add.cuh>
@@ -38,7 +33,12 @@
 #include <raft/linalg/unary_op.cuh>
 #include <raft/matrix/gather.cuh>
 #include <raft/matrix/linewise_op.cuh>
+#include <raft/neighbors/detail/ivf_common.cuh>
+#include <raft/neighbors/detail/ivf_pq_codepacking.cuh>
+#include <raft/neighbors/ivf_list.hpp>
+#include <raft/neighbors/ivf_pq_types.hpp>
 #include <raft/random/rng.cuh>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/stats/histogram.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/device_atomics.cuh>
@@ -46,16 +46,14 @@
 #include <raft/util/pow2_utils.cuh>
 #include <raft/util/vectorized.cuh>
 
-#include <raft/core/resource/device_memory_resource.hpp>
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/managed_memory_resource.hpp>
 
+#include <cuda_fp16.h>
 #include <thrust/extrema.h>
 #include <thrust/scan.h>
 
-#include <cuda_fp16.h>
-
 #include <memory>
 #include <variant>
 
@@ -1363,59 +1361,6 @@ void process_and_fill_codes(raft::resources const& handle,
   RAFT_CUDA_TRY(cudaPeekAtLastError());
 }
 
-/** Update the state of the dependent index members. */
-template <typename IdxT>
-void recompute_internal_state(const raft::resources& res, index<IdxT>& index)
-{
-  auto stream  = resource::get_cuda_stream(res);
-  auto tmp_res = resource::get_workspace_resource(res);
-  rmm::device_uvector<uint32_t> sorted_sizes(index.n_lists(), stream, tmp_res);
-
-  // Actualize the list pointers
-  auto data_ptrs = index.data_ptrs();
-  auto inds_ptrs = index.inds_ptrs();
-  for (uint32_t label = 0; label < index.n_lists(); label++) {
-    auto& list          = index.lists()[label];
-    const auto data_ptr = list ? list->data.data_handle() : nullptr;
-    const auto inds_ptr = list ? list->indices.data_handle() : nullptr;
-    copy(&data_ptrs(label), &data_ptr, 1, stream);
-    copy(&inds_ptrs(label), &inds_ptr, 1, stream);
-  }
-
-  // Sort the cluster sizes in the descending order.
-  int begin_bit             = 0;
-  int end_bit               = sizeof(uint32_t) * 8;
-  size_t cub_workspace_size = 0;
-  cub::DeviceRadixSort::SortKeysDescending(nullptr,
-                                           cub_workspace_size,
-                                           index.list_sizes().data_handle(),
-                                           sorted_sizes.data(),
-                                           index.n_lists(),
-                                           begin_bit,
-                                           end_bit,
-                                           stream);
-  rmm::device_buffer cub_workspace(cub_workspace_size, stream, tmp_res);
-  cub::DeviceRadixSort::SortKeysDescending(cub_workspace.data(),
-                                           cub_workspace_size,
-                                           index.list_sizes().data_handle(),
-                                           sorted_sizes.data(),
-                                           index.n_lists(),
-                                           begin_bit,
-                                           end_bit,
-                                           stream);
-  // copy the results to CPU
-  std::vector<uint32_t> sorted_sizes_host(index.n_lists());
-  copy(sorted_sizes_host.data(), sorted_sizes.data(), index.n_lists(), stream);
-  resource::sync_stream(res);
-
-  // accumulate the sorted cluster sizes
-  auto accum_sorted_sizes = index.accum_sorted_sizes();
-  accum_sorted_sizes(0)   = 0;
-  for (uint32_t label = 0; label < sorted_sizes_host.size(); label++) {
-    accum_sorted_sizes(label + 1) = accum_sorted_sizes(label) + sorted_sizes_host[label];
-  }
-}
-
 /**
  * Helper function: allocate enough space in the list, compute the offset, at which to start
  * writing, and fill-in indices.
@@ -1463,7 +1408,7 @@ void extend_list_with_codes(raft::resources const& res,
   // Pack the data
   pack_list_data<IdxT>(res, index, new_codes, label, offset);
   // Update the pointers and the sizes
-  recompute_internal_state(res, *index);
+  ivf::detail::recompute_internal_state(res, *index);
 }
 
 /**
@@ -1482,7 +1427,7 @@ void extend_list(raft::resources const& res,
   // Encode the data
   encode_list_data<T, IdxT>(res, index, new_vectors, label, offset);
   // Update the pointers and the sizes
-  recompute_internal_state(res, *index);
+  ivf::detail::recompute_internal_state(res, *index);
 }
 
 /**
@@ -1495,7 +1440,7 @@ void erase_list(raft::resources const& res, index<IdxT>* index, uint32_t label)
   uint32_t zero = 0;
   copy(index->list_sizes().data_handle() + label, &zero, 1, resource::get_cuda_stream(res));
   index->lists()[label].reset();
-  recompute_internal_state(res, *index);
+  ivf::detail::recompute_internal_state(res, *index);
 }
 
 /** Copy the state of an index into a new index, but share the list data among the two. */
@@ -1539,7 +1484,7 @@ auto clone(const raft::resources& res, const index<IdxT>& source) -> index<IdxT>
   target.lists() = source.lists();
 
   // Make sure the device pointers point to the new lists
-  recompute_internal_state(res, target);
+  ivf::detail::recompute_internal_state(res, target);
 
   return target;
 }
@@ -1688,7 +1633,7 @@ void extend(raft::resources const& handle,
   }
 
   // Update the pointers and the sizes
-  recompute_internal_state(handle, *index);
+  ivf::detail::recompute_internal_state(handle, *index);
 
   // Recover old cluster sizes: they are used as counters in the fill-codes kernel
   copy(list_sizes, orig_list_sizes.data(), n_clusters, stream);
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_codepacking.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_codepacking.cuh
index 52969dd176..bd03409f66 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_codepacking.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_codepacking.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,9 @@
 
 #pragma once
 
+#include <raft/core/device_mdspan.hpp>
 #include <raft/neighbors/ivf_list.hpp>
 #include <raft/neighbors/ivf_pq_types.hpp>
-
-#include <raft/core/device_mdspan.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/device_atomics.cuh>
 #include <raft/util/integer_utils.hpp>
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-ext.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-ext.cuh
index 37b6efc1eb..5e1a9b46d6 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-ext.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-ext.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,16 @@
 
 #pragma once
 
-#include <cuda_fp16.h>                               // __half
 #include <raft/core/detail/macros.hpp>               // RAFT_WEAK_FUNCTION
 #include <raft/distance/distance_types.hpp>          // raft::distance::DistanceType
 #include <raft/neighbors/detail/ivf_pq_fp_8bit.cuh>  // raft::neighbors::ivf_pq::detail::fp_8bit
 #include <raft/neighbors/ivf_pq_types.hpp>           // raft::neighbors::ivf_pq::codebook_gen
 #include <raft/neighbors/sample_filter_types.hpp>    // none_ivf_sample_filter
 #include <raft/util/raft_explicit.hpp>               // RAFT_EXPLICIT
-#include <rmm/cuda_stream_view.hpp>                  // rmm::cuda_stream_view
+
+#include <rmm/cuda_stream_view.hpp>  // rmm::cuda_stream_view
+
+#include <cuda_fp16.h>  // __half
 
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-inl.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-inl.cuh
index bd88c029e1..462d134b8e 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-inl.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_compute_similarity-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,14 +18,15 @@
 
 #include <raft/distance/distance_types.hpp>  // raft::distance::DistanceType
 #include <raft/matrix/detail/select_warpsort.cuh>  // matrix::detail::select::warpsort::warp_sort_distributed
-#include <raft/neighbors/detail/ivf_pq_dummy_block_sort.cuh>  // dummy_block_sort_t
-#include <raft/neighbors/ivf_pq_types.hpp>                    // codebook_gen
-#include <raft/neighbors/sample_filter_types.hpp>             // none_ivf_sample_filter
-#include <raft/util/cuda_rt_essentials.hpp>                   // RAFT_CUDA_TRY
-#include <raft/util/device_atomics.cuh>                       // raft::atomicMin
-#include <raft/util/pow2_utils.cuh>                           // raft::Pow2
-#include <raft/util/vectorized.cuh>                           // raft::TxN_t
-#include <rmm/cuda_stream_view.hpp>                           // rmm::cuda_stream_view
+#include <raft/neighbors/detail/ivf_common.cuh>    // dummy_block_sort_t
+#include <raft/neighbors/ivf_pq_types.hpp>         // codebook_gen
+#include <raft/neighbors/sample_filter_types.hpp>  // none_ivf_sample_filter
+#include <raft/util/cuda_rt_essentials.hpp>        // RAFT_CUDA_TRY
+#include <raft/util/device_atomics.cuh>            // raft::atomicMin
+#include <raft/util/pow2_utils.cuh>                // raft::Pow2
+#include <raft/util/vectorized.cuh>                // raft::TxN_t
+
+#include <rmm/cuda_stream_view.hpp>  // rmm::cuda_stream_view
 
 namespace raft::neighbors::ivf_pq::detail {
 
@@ -72,8 +73,8 @@ struct pq_block_sort {
 };
 
 template <typename T, typename IdxT>
-struct pq_block_sort<0, T, IdxT> : dummy_block_sort_t<T, IdxT> {
-  using type = dummy_block_sort_t<T, IdxT>;
+struct pq_block_sort<0, T, IdxT> : ivf::detail::dummy_block_sort_t<T, IdxT> {
+  using type = ivf::detail::dummy_block_sort_t<T, IdxT>;
   static auto mem_required(uint32_t) -> size_t { return 0; }
   static auto get_mem_required(uint32_t) { return mem_required; }
 };
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_dummy_block_sort.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_dummy_block_sort.cuh
deleted file mode 100644
index a00b6a50ff..0000000000
--- a/cpp/include/raft/neighbors/detail/ivf_pq_dummy_block_sort.cuh
+++ /dev/null
@@ -1,39 +0,0 @@
-/*
- * Copyright (c) 2023, NVIDIA CORPORATION.
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *     http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-#pragma once
-
-#include <raft/matrix/detail/select_warpsort.cuh>  // matrix::detail::select::warpsort::warp_sort_distributed
-
-/*
- * This header file is a bit of an ugly duckling. The type dummy_block_sort is
- * needed by both ivf_pq_search.cuh and ivf_pq_compute_similarity.cuh.
- *
- * I have decided to move it to it's own header file, which is overkill. Perhaps
- * there is a nicer solution.
- *
- */
-
-namespace raft::neighbors::ivf_pq::detail {
-
-template <typename T, typename IdxT>
-struct dummy_block_sort_t {
-  using queue_t = matrix::detail::select::warpsort::warp_sort_distributed<WarpSize, true, T, IdxT>;
-  template <typename... Args>
-  __device__ dummy_block_sort_t(int k, Args...){};
-};
-
-}  // namespace raft::neighbors::ivf_pq::detail
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_fp_8bit.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_fp_8bit.cuh
index 68c8a513f6..1dea998f9b 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_fp_8bit.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_fp_8bit.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,10 +16,6 @@
 
 #pragma once
 
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <raft/neighbors/ivf_pq_types.hpp>
-
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/logger.hpp>
@@ -32,6 +28,8 @@
 #include <raft/linalg/unary_op.cuh>
 #include <raft/matrix/detail/select_k.cuh>
 #include <raft/matrix/detail/select_warpsort.cuh>
+#include <raft/neighbors/ivf_pq_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/device_atomics.cuh>
 #include <raft/util/device_loads_stores.cuh>
@@ -42,7 +40,6 @@
 #include <rmm/mr/device/per_device_resource.hpp>
 
 #include <cub/cub.cuh>
-
 #include <cuda_fp16.h>
 
 #include <optional>
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_search.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_search.cuh
index d000a1a4d3..d445f909e5 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_search.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_search.cuh
@@ -16,23 +16,15 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/device_properties.hpp>
-#include <raft/spatial/knn/detail/ann_utils.cuh>
-
-#include <raft/neighbors/detail/ivf_pq_compute_similarity.cuh>
-#include <raft/neighbors/detail/ivf_pq_dummy_block_sort.cuh>
-#include <raft/neighbors/detail/ivf_pq_fp_8bit.cuh>
-#include <raft/neighbors/ivf_pq_types.hpp>
-#include <raft/neighbors/sample_filter_types.hpp>
-
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/logger.hpp>
 #include <raft/core/nvtx.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/custom_resource.hpp>
 #include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/core/resource/device_properties.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/gemm.cuh>
@@ -40,6 +32,12 @@
 #include <raft/linalg/unary_op.cuh>
 #include <raft/matrix/detail/select_k.cuh>
 #include <raft/matrix/detail/select_warpsort.cuh>
+#include <raft/neighbors/detail/ivf_common.cuh>
+#include <raft/neighbors/detail/ivf_pq_compute_similarity.cuh>
+#include <raft/neighbors/detail/ivf_pq_fp_8bit.cuh>
+#include <raft/neighbors/ivf_pq_types.hpp>
+#include <raft/neighbors/sample_filter_types.hpp>
+#include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cache.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/device_atomics.cuh>
@@ -51,7 +49,6 @@
 #include <rmm/mr/device/per_device_resource.hpp>
 
 #include <cub/cub.cuh>
-
 #include <cuda_fp16.h>
 
 #include <optional>
@@ -171,217 +168,6 @@ void select_clusters(raft::resources const& handle,
                                             true);
 }
 
-/**
- * For each query, we calculate a cumulative sum of the cluster sizes that we probe, and return that
- * in chunk_indices. Essentially this is a segmented inclusive scan of the cluster sizes. The total
- * number of samples per query (sum of the cluster sizes that we probe) is returned in n_samples.
- */
-template <int BlockDim>
-__launch_bounds__(BlockDim) RAFT_KERNEL
-  calc_chunk_indices_kernel(uint32_t n_probes,
-                            const uint32_t* cluster_sizes,      // [n_clusters]
-                            const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
-                            uint32_t* chunk_indices,            // [n_queries, n_probes]
-                            uint32_t* n_samples                 // [n_queries]
-  )
-{
-  using block_scan = cub::BlockScan<uint32_t, BlockDim>;
-  __shared__ typename block_scan::TempStorage shm;
-
-  // locate the query data
-  clusters_to_probe += n_probes * blockIdx.x;
-  chunk_indices += n_probes * blockIdx.x;
-
-  // block scan
-  const uint32_t n_probes_aligned = Pow2<BlockDim>::roundUp(n_probes);
-  uint32_t total                  = 0;
-  for (uint32_t probe_ix = threadIdx.x; probe_ix < n_probes_aligned; probe_ix += BlockDim) {
-    auto label = probe_ix < n_probes ? clusters_to_probe[probe_ix] : 0u;
-    auto chunk = probe_ix < n_probes ? cluster_sizes[label] : 0u;
-    if (threadIdx.x == 0) { chunk += total; }
-    block_scan(shm).InclusiveSum(chunk, chunk, total);
-    __syncthreads();
-    if (probe_ix < n_probes) { chunk_indices[probe_ix] = chunk; }
-  }
-  // save the total size
-  if (threadIdx.x == 0) { n_samples[blockIdx.x] = total; }
-}
-
-struct calc_chunk_indices {
- public:
-  struct configured {
-    void* kernel;
-    dim3 block_dim;
-    dim3 grid_dim;
-    uint32_t n_probes;
-
-    inline void operator()(const uint32_t* cluster_sizes,
-                           const uint32_t* clusters_to_probe,
-                           uint32_t* chunk_indices,
-                           uint32_t* n_samples,
-                           rmm::cuda_stream_view stream)
-    {
-      void* args[] =  // NOLINT
-        {&n_probes, &cluster_sizes, &clusters_to_probe, &chunk_indices, &n_samples};
-      RAFT_CUDA_TRY(cudaLaunchKernel(kernel, grid_dim, block_dim, args, 0, stream));
-    }
-  };
-
-  static inline auto configure(uint32_t n_probes, uint32_t n_queries) -> configured
-  {
-    return try_block_dim<1024>(n_probes, n_queries);
-  }
-
- private:
-  template <int BlockDim>
-  static auto try_block_dim(uint32_t n_probes, uint32_t n_queries) -> configured
-  {
-    if constexpr (BlockDim >= WarpSize * 2) {
-      if (BlockDim >= n_probes * 2) { return try_block_dim<(BlockDim / 2)>(n_probes, n_queries); }
-    }
-    return {reinterpret_cast<void*>(calc_chunk_indices_kernel<BlockDim>),
-            dim3(BlockDim, 1, 1),
-            dim3(n_queries, 1, 1),
-            n_probes};
-  }
-};
-
-/**
- * Look up the chunk id corresponding to the sample index.
- *
- * Each query vector was compared to all the vectors from n_probes clusters, and sample_ix is an
- * ordered number of one of such vectors. This function looks up to which chunk it belongs,
- * and returns the index within the chunk (which is also an index within a cluster).
- *
- * @param[inout] sample_ix
- *   input: the offset of the sample in the batch;
- *   output: the offset inside the chunk (probe) / selected cluster.
- * @param[in] n_probes number of probes
- * @param[in] chunk_indices offsets of the chunks within the batch [n_probes]
- * @return chunk index (== n_probes when the input index is not in the valid range,
- *    which can happen if there is not enough data to output in the selected clusters).
- */
-__device__ inline auto find_chunk_ix(uint32_t& sample_ix,  // NOLINT
-                                     uint32_t n_probes,
-                                     const uint32_t* chunk_indices) -> uint32_t
-{
-  uint32_t ix_min = 0;
-  uint32_t ix_max = n_probes;
-  do {
-    uint32_t i = (ix_min + ix_max) / 2;
-    if (chunk_indices[i] <= sample_ix) {
-      ix_min = i + 1;
-    } else {
-      ix_max = i;
-    }
-  } while (ix_min < ix_max);
-  if (ix_min > 0) { sample_ix -= chunk_indices[ix_min - 1]; }
-  return ix_min;
-}
-
-template <int BlockDim, typename IdxT>
-__launch_bounds__(BlockDim) RAFT_KERNEL
-  postprocess_neighbors_kernel(IdxT* neighbors_out,                // [n_queries, topk]
-                               const uint32_t* neighbors_in,       // [n_queries, topk]
-                               const IdxT* const* db_indices,      // [n_clusters][..]
-                               const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
-                               const uint32_t* chunk_indices,      // [n_queries, n_probes]
-                               uint32_t n_queries,
-                               uint32_t n_probes,
-                               uint32_t topk)
-{
-  const uint64_t i        = threadIdx.x + BlockDim * uint64_t(blockIdx.x);
-  const uint32_t query_ix = i / uint64_t(topk);
-  if (query_ix >= n_queries) { return; }
-  const uint32_t k = i % uint64_t(topk);
-  neighbors_in += query_ix * topk;
-  neighbors_out += query_ix * topk;
-  chunk_indices += query_ix * n_probes;
-  clusters_to_probe += query_ix * n_probes;
-  uint32_t data_ix        = neighbors_in[k];
-  const uint32_t chunk_ix = find_chunk_ix(data_ix, n_probes, chunk_indices);
-  const bool valid        = chunk_ix < n_probes;
-  neighbors_out[k] =
-    valid ? db_indices[clusters_to_probe[chunk_ix]][data_ix] : ivf_pq::kOutOfBoundsRecord<IdxT>;
-}
-
-/**
- * Transform found sample indices into the corresponding database indices
- * (as stored in index.indices()).
- * The sample indices are the record indices as they appear in the database view formed by the
- * probed clusters / defined by the `chunk_indices`.
- * We assume the searched sample sizes (for a single query) fit into `uint32_t`.
- */
-template <typename IdxT>
-void postprocess_neighbors(IdxT* neighbors_out,                // [n_queries, topk]
-                           const uint32_t* neighbors_in,       // [n_queries, topk]
-                           const IdxT* const* db_indices,      // [n_clusters][..]
-                           const uint32_t* clusters_to_probe,  // [n_queries, n_probes]
-                           const uint32_t* chunk_indices,      // [n_queries, n_probes]
-                           uint32_t n_queries,
-                           uint32_t n_probes,
-                           uint32_t topk,
-                           rmm::cuda_stream_view stream)
-{
-  constexpr int kPNThreads = 256;
-  const int pn_blocks      = raft::div_rounding_up_unsafe<size_t>(n_queries * topk, kPNThreads);
-  postprocess_neighbors_kernel<kPNThreads, IdxT>
-    <<<pn_blocks, kPNThreads, 0, stream>>>(neighbors_out,
-                                           neighbors_in,
-                                           db_indices,
-                                           clusters_to_probe,
-                                           chunk_indices,
-                                           n_queries,
-                                           n_probes,
-                                           topk);
-}
-
-/**
- * Post-process the scores depending on the metric type;
- * translate the element type if necessary.
- */
-template <typename ScoreT>
-void postprocess_distances(float* out,        // [n_queries, topk]
-                           const ScoreT* in,  // [n_queries, topk]
-                           distance::DistanceType metric,
-                           uint32_t n_queries,
-                           uint32_t topk,
-                           float scaling_factor,
-                           rmm::cuda_stream_view stream)
-{
-  size_t len = size_t(n_queries) * size_t(topk);
-  switch (metric) {
-    case distance::DistanceType::L2Unexpanded:
-    case distance::DistanceType::L2Expanded: {
-      linalg::unaryOp(out,
-                      in,
-                      len,
-                      raft::compose_op(raft::mul_const_op<float>{scaling_factor * scaling_factor},
-                                       raft::cast_op<float>{}),
-                      stream);
-    } break;
-    case distance::DistanceType::L2SqrtUnexpanded:
-    case distance::DistanceType::L2SqrtExpanded: {
-      linalg::unaryOp(
-        out,
-        in,
-        len,
-        raft::compose_op{
-          raft::mul_const_op<float>{scaling_factor}, raft::sqrt_op{}, raft::cast_op<float>{}},
-        stream);
-    } break;
-    case distance::DistanceType::InnerProduct: {
-      linalg::unaryOp(out,
-                      in,
-                      len,
-                      raft::compose_op(raft::mul_const_op<float>{-scaling_factor * scaling_factor},
-                                       raft::cast_op<float>{}),
-                      stream);
-    } break;
-    default: RAFT_FAIL("Unexpected metric.");
-  }
-}
-
 /**
  * An approximation to the number of times each cluster appears in a batched sample.
  *
@@ -520,11 +306,11 @@ void ivfpq_search_worker(raft::resources const& handle,
     neighbors_uint32 = neighbors_uint32_buf.data();
   }
 
-  calc_chunk_indices::configure(n_probes, n_queries)(index.list_sizes().data_handle(),
-                                                     clusters_to_probe,
-                                                     chunk_index.data(),
-                                                     num_samples.data(),
-                                                     stream);
+  ivf::detail::calc_chunk_indices::configure(n_probes, n_queries)(index.list_sizes().data_handle(),
+                                                                  clusters_to_probe,
+                                                                  chunk_index.data(),
+                                                                  num_samples.data(),
+                                                                  stream);
 
   auto coresidency = expected_probe_coresidency(index.n_lists(), n_probes, n_queries);
 
@@ -621,9 +407,10 @@ void ivfpq_search_worker(raft::resources const& handle,
   if (manage_local_topk) {
     query_kths_buf.emplace(
       make_device_mdarray<float>(handle, mr, make_extents<uint32_t>(n_queries)));
-    linalg::map(handle,
-                query_kths_buf->view(),
-                raft::const_op<float>{dummy_block_sort_t<ScoreT, IdxT>::queue_t::kDummy});
+    linalg::map(
+      handle,
+      query_kths_buf->view(),
+      raft::const_op<float>{ivf::detail::dummy_block_sort_t<ScoreT, IdxT>::queue_t::kDummy});
     query_kths = query_kths_buf->data_handle();
   }
   compute_similarity_run(search_instance,
@@ -663,17 +450,17 @@ void ivfpq_search_worker(raft::resources const& handle,
                                              true);
 
   // Postprocessing
-  postprocess_distances(
-    distances, topk_dists.data(), index.metric(), n_queries, topK, scaling_factor, stream);
-  postprocess_neighbors(neighbors,
-                        neighbors_uint32,
-                        index.inds_ptrs().data_handle(),
-                        clusters_to_probe,
-                        chunk_index.data(),
-                        n_queries,
-                        n_probes,
-                        topK,
-                        stream);
+  ivf::detail::postprocess_distances(
+    distances, topk_dists.data(), index.metric(), n_queries, topK, scaling_factor, true, stream);
+  ivf::detail::postprocess_neighbors(neighbors,
+                                     neighbors_uint32,
+                                     index.inds_ptrs().data_handle(),
+                                     clusters_to_probe,
+                                     chunk_index.data(),
+                                     n_queries,
+                                     n_probes,
+                                     topK,
+                                     stream);
 }
 
 /**
diff --git a/cpp/include/raft/neighbors/detail/ivf_pq_serialize.cuh b/cpp/include/raft/neighbors/detail/ivf_pq_serialize.cuh
index 038bf8d7cc..4428fa370b 100644
--- a/cpp/include/raft/neighbors/detail/ivf_pq_serialize.cuh
+++ b/cpp/include/raft/neighbors/detail/ivf_pq_serialize.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,14 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/neighbors/detail/ivf_pq_build.cuh>
-#include <raft/neighbors/ivf_list.hpp>
-#include <raft/neighbors/ivf_pq_types.hpp>
-
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/core/serialize.hpp>
+#include <raft/neighbors/detail/ivf_pq_build.cuh>
+#include <raft/neighbors/ivf_list.hpp>
+#include <raft/neighbors/ivf_pq_types.hpp>
 
 #include <fstream>
 #include <memory>
@@ -160,7 +159,7 @@ auto deserialize(raft::resources const& handle_, std::istream& is) -> index<IdxT
 
   resource::sync_stream(handle_);
 
-  recompute_internal_state(handle_, index);
+  ivf::detail::recompute_internal_state(handle_, index);
 
   return index;
 }
diff --git a/cpp/include/raft/neighbors/detail/knn_merge_parts.cuh b/cpp/include/raft/neighbors/detail/knn_merge_parts.cuh
index c8ff03741c..0395e86e43 100644
--- a/cpp/include/raft/neighbors/detail/knn_merge_parts.cuh
+++ b/cpp/include/raft/neighbors/detail/knn_merge_parts.cuh
@@ -16,12 +16,12 @@
 
 #pragma once
 
+#include <raft/neighbors/detail/faiss_select/DistanceUtils.h>
+#include <raft/neighbors/detail/faiss_select/Select.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
 #include <cstdint>
-#include <raft/neighbors/detail/faiss_select/DistanceUtils.h>
-#include <raft/neighbors/detail/faiss_select/Select.cuh>
 
 namespace raft::neighbors::detail {
 
diff --git a/cpp/include/raft/neighbors/detail/nn_descent.cuh b/cpp/include/raft/neighbors/detail/nn_descent.cuh
index f624a6015b..cd3c6f3947 100644
--- a/cpp/include/raft/neighbors/detail/nn_descent.cuh
+++ b/cpp/include/raft/neighbors/detail/nn_descent.cuh
@@ -16,23 +16,6 @@
 
 #pragma once
 
-#include <cuda_runtime.h>
-#include <mma.h>
-#include <omp.h>
-
-#include <cub/cub.cuh>
-#include <limits>
-#include <queue>
-
-#include <random>
-#include <rmm/device_uvector.hpp>
-
-#include <thrust/execution_policy.h>
-#include <thrust/fill.h>
-#include <thrust/host_vector.h>
-#include <thrust/mr/allocator.h>
-#include <thrust/mr/device_memory_resource.h>
-
 #include "../nn_descent_types.hpp"
 
 #include <raft/core/device_mdarray.hpp>
@@ -48,6 +31,23 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/pow2_utils.cuh>
 
+#include <rmm/device_uvector.hpp>
+
+#include <cub/cub.cuh>
+#include <cuda_runtime.h>
+#include <thrust/execution_policy.h>
+#include <thrust/fill.h>
+#include <thrust/host_vector.h>
+#include <thrust/mr/allocator.h>
+#include <thrust/mr/device_memory_resource.h>
+
+#include <mma.h>
+#include <omp.h>
+
+#include <limits>
+#include <queue>
+#include <random>
+
 namespace raft::neighbors::experimental::nn_descent::detail {
 
 using pinned_memory_resource = thrust::universal_host_pinned_memory_resource;
diff --git a/cpp/include/raft/neighbors/detail/refine_device.cuh b/cpp/include/raft/neighbors/detail/refine_device.cuh
index 5c9f1459e7..e76e52657b 100644
--- a/cpp/include/raft/neighbors/detail/refine_device.cuh
+++ b/cpp/include/raft/neighbors/detail/refine_device.cuh
@@ -50,6 +50,7 @@ void refine_device(raft::resources const& handle,
   matrix_idx dim          = dataset.extent(1);
   uint32_t k              = static_cast<uint32_t>(indices.extent(1));
 
+  // TODO: this restriction could be lifted with some effort
   RAFT_EXPECTS(k <= raft::matrix::detail::select::warpsort::kMaxCapacity,
                "k must be less than topk::kMaxCapacity (%d).",
                raft::matrix::detail::select::warpsort::kMaxCapacity);
@@ -98,6 +99,8 @@ void refine_device(raft::resources const& handle,
            refinement_index.metric(),
            1,
            k,
+           0,
+           nullptr,
            raft::distance::is_min_close(metric),
            raft::neighbors::filtering::none_ivf_sample_filter(),
            indices.data_handle(),
diff --git a/cpp/include/raft/neighbors/detail/refine_host-ext.hpp b/cpp/include/raft/neighbors/detail/refine_host-ext.hpp
index a2f75d0f8b..69d2bd29b2 100644
--- a/cpp/include/raft/neighbors/detail/refine_host-ext.hpp
+++ b/cpp/include/raft/neighbors/detail/refine_host-ext.hpp
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <cstdint>  // int64_t
-
 #include <raft/core/detail/macros.hpp>       // _RAFT_HAS_CUDA
 #include <raft/core/host_mdspan.hpp>         // raft::host_matrix_view
 #include <raft/distance/distance_types.hpp>  // raft::distance::DistanceType
 #include <raft/util/raft_explicit.hpp>       // RAFT_EXPLICIT
 
+#include <cstdint>  // int64_t
+
 #if defined(_RAFT_HAS_CUDA)
 #include <cuda_fp16.h>
 #endif
diff --git a/cpp/include/raft/neighbors/detail/refine_host-inl.hpp b/cpp/include/raft/neighbors/detail/refine_host-inl.hpp
index a54525f3e6..9aff451dfc 100644
--- a/cpp/include/raft/neighbors/detail/refine_host-inl.hpp
+++ b/cpp/include/raft/neighbors/detail/refine_host-inl.hpp
@@ -21,9 +21,10 @@
 #include <raft/neighbors/detail/refine_common.hpp>
 #include <raft/util/integer_utils.hpp>
 
-#include <algorithm>
 #include <omp.h>
 
+#include <algorithm>
+
 namespace raft::neighbors::detail {
 
 template <typename DC, typename IdxT, typename DataT, typename DistanceT, typename ExtentsT>
diff --git a/cpp/include/raft/neighbors/hnsw.hpp b/cpp/include/raft/neighbors/hnsw.hpp
index dceb98c5aa..964c3ffacd 100644
--- a/cpp/include/raft/neighbors/hnsw.hpp
+++ b/cpp/include/raft/neighbors/hnsw.hpp
@@ -19,13 +19,13 @@
 #include "detail/hnsw.hpp"
 #include "hnsw.hpp"
 
-#include <cstddef>
-
-#include <cstdint>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/neighbors/cagra_types.hpp>
 
+#include <cstddef>
+#include <cstdint>
+
 namespace raft::neighbors::hnsw {
 
 /**
diff --git a/cpp/include/raft/neighbors/hnsw_serialize.hpp b/cpp/include/raft/neighbors/hnsw_serialize.hpp
index 45819c8fb5..f7450bb3d6 100644
--- a/cpp/include/raft/neighbors/hnsw_serialize.hpp
+++ b/cpp/include/raft/neighbors/hnsw_serialize.hpp
@@ -18,9 +18,9 @@
 
 #include "detail/hnsw_serialize.hpp"
 #include "hnsw_types.hpp"
-#include <raft/distance/distance_types.hpp>
 
 #include <raft/core/resources.hpp>
+#include <raft/distance/distance_types.hpp>
 
 namespace raft::neighbors::hnsw {
 
diff --git a/cpp/include/raft/neighbors/hnsw_types.hpp b/cpp/include/raft/neighbors/hnsw_types.hpp
index aa4cefbc30..645a0903b7 100644
--- a/cpp/include/raft/neighbors/hnsw_types.hpp
+++ b/cpp/include/raft/neighbors/hnsw_types.hpp
@@ -17,11 +17,13 @@
 #pragma once
 
 #include "ann_types.hpp"
-#include <memory>
+
 #include <raft/distance/distance_types.hpp>
 
-#include <cstdint>
 #include <sys/types.h>
+
+#include <cstdint>
+#include <memory>
 #include <type_traits>
 
 namespace raft::neighbors::hnsw {
diff --git a/cpp/include/raft/neighbors/ivf_flat-ext.cuh b/cpp/include/raft/neighbors/ivf_flat-ext.cuh
index 063105cf46..a1783dfcfd 100644
--- a/cpp/include/raft/neighbors/ivf_flat-ext.cuh
+++ b/cpp/include/raft/neighbors/ivf_flat-ext.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,16 @@
 
 #pragma once
 
-#include <cstdint>  // int64_t
-
 #include <raft/core/device_mdspan.hpp>  // raft::device_matrix_view
 #include <raft/core/resources.hpp>      // raft::resources
 #include <raft/neighbors/ivf_flat_serialize.cuh>
-#include <raft/neighbors/ivf_flat_types.hpp>      // raft::neighbors::ivf_flat::index
-#include <raft/util/raft_explicit.hpp>            // RAFT_EXPLICIT
+#include <raft/neighbors/ivf_flat_types.hpp>  // raft::neighbors::ivf_flat::index
+#include <raft/util/raft_explicit.hpp>        // RAFT_EXPLICIT
+
 #include <rmm/mr/device/per_device_resource.hpp>  // rmm::mr::device_memory_resource
 
+#include <cstdint>  // int64_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::ivf_flat {
diff --git a/cpp/include/raft/neighbors/ivf_flat-inl.cuh b/cpp/include/raft/neighbors/ivf_flat-inl.cuh
index 13c8e3d86d..ed1d320795 100644
--- a/cpp/include/raft/neighbors/ivf_flat-inl.cuh
+++ b/cpp/include/raft/neighbors/ivf_flat-inl.cuh
@@ -16,14 +16,13 @@
 
 #pragma once
 
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/neighbors/detail/ivf_flat_build.cuh>
 #include <raft/neighbors/detail/ivf_flat_search.cuh>
 #include <raft/neighbors/ivf_flat_serialize.cuh>
 #include <raft/neighbors/ivf_flat_types.hpp>
 
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resources.hpp>
-
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/mr/device/per_device_resource.hpp>
 
diff --git a/cpp/include/raft/neighbors/ivf_flat_helpers.cuh b/cpp/include/raft/neighbors/ivf_flat_helpers.cuh
index 7a05c9991c..883e72c839 100644
--- a/cpp/include/raft/neighbors/ivf_flat_helpers.cuh
+++ b/cpp/include/raft/neighbors/ivf_flat_helpers.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,10 @@
 
 #pragma once
 
-#include <raft/neighbors/detail/ivf_flat_build.cuh>
-#include <raft/neighbors/ivf_flat_types.hpp>
-
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resources.hpp>
-
+#include <raft/neighbors/detail/ivf_flat_build.cuh>
+#include <raft/neighbors/ivf_flat_types.hpp>
 #include <raft/spatial/knn/detail/ann_utils.cuh>
 
 namespace raft::neighbors::ivf_flat::helpers {
@@ -137,9 +135,42 @@ void reset_index(const raft::resources& res, index<T, IdxT>* index)
 {
   auto stream = resource::get_cuda_stream(res);
 
+  utils::memzero(
+    index->accum_sorted_sizes().data_handle(), index->accum_sorted_sizes().size(), stream);
   utils::memzero(index->list_sizes().data_handle(), index->list_sizes().size(), stream);
   utils::memzero(index->data_ptrs().data_handle(), index->data_ptrs().size(), stream);
   utils::memzero(index->inds_ptrs().data_handle(), index->inds_ptrs().size(), stream);
 }
+
+/**
+ * @brief Helper exposing the re-computation of list sizes and related arrays if IVF lists have been
+ * modified.
+ *
+ * Usage example:
+ * @code{.cpp}
+ *   using namespace raft::neighbors;
+ *   raft::resources res;
+ *   // use default index parameters
+ *   ivf_flat::index_params index_params;
+ *   // initialize an empty index
+ *   ivf_flat::index<int64_t> index(res, index_params, D);
+ *   ivf_flat::helpers::reset_index(res, &index);
+ *   // recompute the internal state of the index
+ *   ivf_flat::helpers::recompute_internal_state(res, &index);
+ * @endcode
+ *
+ * @tparam T
+ * @tparam IdxT
+ *
+ * @param[in] res raft resource
+ * @param[inout] index pointer to IVF-FLAT index
+ */
+template <typename T, typename IdxT>
+void recompute_internal_state(const raft::resources& res, index<T, IdxT>* index)
+{
+  auto& list = index->lists()[0];
+  ivf::detail::recompute_internal_state(res, *index);
+}
+
 /** @} */
 }  // namespace raft::neighbors::ivf_flat::helpers
diff --git a/cpp/include/raft/neighbors/ivf_flat_types.hpp b/cpp/include/raft/neighbors/ivf_flat_types.hpp
index 180fe2e21b..0908e3b0b0 100644
--- a/cpp/include/raft/neighbors/ivf_flat_types.hpp
+++ b/cpp/include/raft/neighbors/ivf_flat_types.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,21 +17,19 @@
 #pragma once
 
 #include "ann_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/ivf_list_types.hpp>
 #include <raft/util/integer_utils.hpp>
 
-#include <thrust/reduce.h>
-
 #include <algorithm>  // std::max
 #include <memory>
 #include <optional>
@@ -222,8 +220,31 @@ struct index : ann::index {
     }
   }
 
+  /**
+   * Accumulated list sizes, sorted in descending order [n_lists + 1].
+   * The last value contains the total length of the index.
+   * The value at index zero is always zero.
+   *
+   * That is, the content of this span is as if the `list_sizes` was sorted and then accumulated.
+   *
+   * This span is used during search to estimate the maximum size of the workspace.
+   */
+  inline auto accum_sorted_sizes() noexcept -> host_vector_view<IdxT, uint32_t>
+  {
+    return accum_sorted_sizes_.view();
+  }
+  [[nodiscard]] inline auto accum_sorted_sizes() const noexcept
+    -> host_vector_view<const IdxT, uint32_t>
+  {
+    return accum_sorted_sizes_.view();
+  }
+
   /** Total length of the index. */
-  [[nodiscard]] constexpr inline auto size() const noexcept -> IdxT { return total_size_; }
+  [[nodiscard]] constexpr inline auto size() const noexcept -> IdxT
+  {
+    return accum_sorted_sizes()(n_lists());
+  }
+
   /** Dimensionality of the data. */
   [[nodiscard]] constexpr inline auto dim() const noexcept -> uint32_t
   {
@@ -257,9 +278,10 @@ struct index : ann::index {
       list_sizes_{make_device_vector<uint32_t, uint32_t>(res, n_lists)},
       data_ptrs_{make_device_vector<T*, uint32_t>(res, n_lists)},
       inds_ptrs_{make_device_vector<IdxT*, uint32_t>(res, n_lists)},
-      total_size_{0}
+      accum_sorted_sizes_{make_host_vector<IdxT, uint32_t>(n_lists + 1)}
   {
     check_consistency();
+    accum_sorted_sizes_(n_lists) = 0;
   }
 
   /** Construct an empty index. It needs to be trained and then populated. */
@@ -298,33 +320,6 @@ struct index : ann::index {
     return conservative_memory_allocation_;
   }
 
-  /**
-   * Update the state of the dependent index members.
-   */
-  void recompute_internal_state(raft::resources const& res)
-  {
-    auto stream = resource::get_cuda_stream(res);
-
-    // Actualize the list pointers
-    auto this_lists     = lists();
-    auto this_data_ptrs = data_ptrs();
-    auto this_inds_ptrs = inds_ptrs();
-    for (uint32_t label = 0; label < this_lists.size(); label++) {
-      auto& list          = this_lists[label];
-      const auto data_ptr = list ? list->data.data_handle() : nullptr;
-      const auto inds_ptr = list ? list->indices.data_handle() : nullptr;
-      copy(&this_data_ptrs(label), &data_ptr, 1, stream);
-      copy(&this_inds_ptrs(label), &inds_ptr, 1, stream);
-    }
-    auto this_list_sizes = list_sizes().data_handle();
-    total_size_          = thrust::reduce(resource::get_thrust_policy(res),
-                                 this_list_sizes,
-                                 this_list_sizes + this_lists.size(),
-                                 0,
-                                 raft::add_op{});
-    check_consistency();
-  }
-
   void allocate_center_norms(raft::resources const& res)
   {
     switch (metric_) {
@@ -349,6 +344,20 @@ struct index : ann::index {
     return lists_;
   }
 
+  /** Throw an error if the index content is inconsistent. */
+  void check_consistency()
+  {
+    auto n_lists = lists_.size();
+    RAFT_EXPECTS(dim() % veclen_ == 0, "dimensionality is not a multiple of the veclen");
+    RAFT_EXPECTS(list_sizes_.extent(0) == n_lists, "inconsistent list size");
+    RAFT_EXPECTS(data_ptrs_.extent(0) == n_lists, "inconsistent list size");
+    RAFT_EXPECTS(inds_ptrs_.extent(0) == n_lists, "inconsistent list size");
+    RAFT_EXPECTS(                                       //
+      (centers_.extent(0) == list_sizes_.extent(0)) &&  //
+        (!center_norms_.has_value() || centers_.extent(0) == center_norms_->extent(0)),
+      "inconsistent number of lists (clusters)");
+  }
+
  private:
   /**
    * TODO: in theory, we can lift this to the template parameter and keep it at hardware maximum
@@ -366,21 +375,7 @@ struct index : ann::index {
   // Computed members
   device_vector<T*, uint32_t> data_ptrs_;
   device_vector<IdxT*, uint32_t> inds_ptrs_;
-  IdxT total_size_;
-
-  /** Throw an error if the index content is inconsistent. */
-  void check_consistency()
-  {
-    auto n_lists = lists_.size();
-    RAFT_EXPECTS(dim() % veclen_ == 0, "dimensionality is not a multiple of the veclen");
-    RAFT_EXPECTS(list_sizes_.extent(0) == n_lists, "inconsistent list size");
-    RAFT_EXPECTS(data_ptrs_.extent(0) == n_lists, "inconsistent list size");
-    RAFT_EXPECTS(inds_ptrs_.extent(0) == n_lists, "inconsistent list size");
-    RAFT_EXPECTS(                                       //
-      (centers_.extent(0) == list_sizes_.extent(0)) &&  //
-        (!center_norms_.has_value() || centers_.extent(0) == center_norms_->extent(0)),
-      "inconsistent number of lists (clusters)");
-  }
+  host_vector<IdxT, uint32_t> accum_sorted_sizes_;
 
   static auto calculate_veclen(uint32_t dim) -> uint32_t
   {
diff --git a/cpp/include/raft/neighbors/ivf_list.hpp b/cpp/include/raft/neighbors/ivf_list.hpp
index ad06a3ee71..08879ed059 100644
--- a/cpp/include/raft/neighbors/ivf_list.hpp
+++ b/cpp/include/raft/neighbors/ivf_list.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,15 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-#include <raft/neighbors/ivf_list_types.hpp>
-
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/core/serialize.hpp>
+#include <raft/neighbors/ivf_list_types.hpp>
 #include <raft/util/integer_utils.hpp>
 
 #include <thrust/fill.h>
diff --git a/cpp/include/raft/neighbors/ivf_pq-ext.cuh b/cpp/include/raft/neighbors/ivf_pq-ext.cuh
index 188219dfa2..160a2753a5 100644
--- a/cpp/include/raft/neighbors/ivf_pq-ext.cuh
+++ b/cpp/include/raft/neighbors/ivf_pq-ext.cuh
@@ -16,14 +16,15 @@
 
 #pragma once
 
-#include <cstdint>  // int64_t
+#include <raft/core/device_mdspan.hpp>      // raft::device_matrix_view
+#include <raft/core/resources.hpp>          // raft::resources
+#include <raft/neighbors/ivf_pq_types.hpp>  // raft::neighbors::ivf_pq::index
+#include <raft/util/raft_explicit.hpp>      // RAFT_EXPLICIT
 
-#include <raft/core/device_mdspan.hpp>            // raft::device_matrix_view
-#include <raft/core/resources.hpp>                // raft::resources
-#include <raft/neighbors/ivf_pq_types.hpp>        // raft::neighbors::ivf_pq::index
-#include <raft/util/raft_explicit.hpp>            // RAFT_EXPLICIT
 #include <rmm/mr/device/per_device_resource.hpp>  // rmm::mr::device_memory_resource
 
+#include <cstdint>  // int64_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors::ivf_pq {
diff --git a/cpp/include/raft/neighbors/ivf_pq-inl.cuh b/cpp/include/raft/neighbors/ivf_pq-inl.cuh
index d14456d6f6..a893153e1a 100644
--- a/cpp/include/raft/neighbors/ivf_pq-inl.cuh
+++ b/cpp/include/raft/neighbors/ivf_pq-inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,14 @@
 
 #pragma once
 
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/device_memory_resource.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/neighbors/detail/ivf_pq_build.cuh>
 #include <raft/neighbors/detail/ivf_pq_search.cuh>
 #include <raft/neighbors/ivf_pq_serialize.cuh>
 #include <raft/neighbors/ivf_pq_types.hpp>
 
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resource/device_memory_resource.hpp>
-#include <raft/core/resources.hpp>
-
 #include <rmm/mr/device/device_memory_resource.hpp>
 
 #include <memory>  // shared_ptr
diff --git a/cpp/include/raft/neighbors/ivf_pq_helpers.cuh b/cpp/include/raft/neighbors/ivf_pq_helpers.cuh
index fec31f1c61..8bd3ab2591 100644
--- a/cpp/include/raft/neighbors/ivf_pq_helpers.cuh
+++ b/cpp/include/raft/neighbors/ivf_pq_helpers.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,15 @@
 
 #pragma once
 
-#include <cstdio>
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/neighbors/detail/ivf_pq_build.cuh>
 #include <raft/neighbors/ivf_pq_types.hpp>
-
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/resources.hpp>
-
 #include <raft/spatial/knn/detail/ann_utils.cuh>
 
+#include <cstdio>
+
 namespace raft::neighbors::ivf_pq::helpers {
 using namespace raft::spatial::knn::detail;  // NOLINT
 /**
@@ -733,7 +732,7 @@ void set_centers(raft::resources const& res,
  *   ivf::resize_list(res, list, spec, new_size, 0);
  *   raft::update_device(index.list_sizes(), &new_size, 1, stream);
  *   // recompute the internal state of the index
- *   ivf_pq::recompute_internal_state(res, &index);
+ *   ivf_pq::helpers::recompute_internal_state(res, &index);
  * @endcode
  *
  * @tparam IdxT
@@ -745,7 +744,7 @@ template <typename IdxT>
 void recompute_internal_state(const raft::resources& res, index<IdxT>* index)
 {
   auto& list = index->lists()[0];
-  ivf_pq::detail::recompute_internal_state(res, *index);
+  ivf::detail::recompute_internal_state(res, *index);
 }
 
 /**
diff --git a/cpp/include/raft/neighbors/ivf_pq_types.hpp b/cpp/include/raft/neighbors/ivf_pq_types.hpp
index 3e6cc443b3..81e2886b18 100644
--- a/cpp/include/raft/neighbors/ivf_pq_types.hpp
+++ b/cpp/include/raft/neighbors/ivf_pq_types.hpp
@@ -16,15 +16,14 @@
 
 #pragma once
 
-#include <raft/neighbors/ann_types.hpp>
-#include <raft/neighbors/ivf_list_types.hpp>
-
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
+#include <raft/neighbors/ann_types.hpp>
+#include <raft/neighbors/ivf_list_types.hpp>
 #include <raft/util/integer_utils.hpp>
 
 #include <thrust/fill.h>
diff --git a/cpp/include/raft/neighbors/nn_descent_types.hpp b/cpp/include/raft/neighbors/nn_descent_types.hpp
index fd1df2965e..e1fc96878a 100644
--- a/cpp/include/raft/neighbors/nn_descent_types.hpp
+++ b/cpp/include/raft/neighbors/nn_descent_types.hpp
@@ -17,11 +17,11 @@
 #pragma once
 
 #include "ann_types.hpp"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/mdspan_types.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
diff --git a/cpp/include/raft/neighbors/refine-ext.cuh b/cpp/include/raft/neighbors/refine-ext.cuh
index 3ccd3891b7..fc57494b22 100644
--- a/cpp/include/raft/neighbors/refine-ext.cuh
+++ b/cpp/include/raft/neighbors/refine-ext.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,14 @@
 
 #pragma once
 
-#include <cstdint>  // int64_t
-
 #include <raft/core/device_mdspan.hpp>       // raft::device_matrix_view
 #include <raft/core/host_mdspan.hpp>         // // raft::host_matrix_view
 #include <raft/core/resources.hpp>           // raft::resources
 #include <raft/distance/distance_types.hpp>  // raft::distance::DistanceType
 #include <raft/util/raft_explicit.hpp>       // RAFT_EXPLICIT
 
+#include <cstdint>  // int64_t
+
 #ifdef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 namespace raft::neighbors {
diff --git a/cpp/include/raft/neighbors/sample_filter.cuh b/cpp/include/raft/neighbors/sample_filter.cuh
index 5864590034..5e9790cf24 100644
--- a/cpp/include/raft/neighbors/sample_filter.cuh
+++ b/cpp/include/raft/neighbors/sample_filter.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,11 @@
 
 #pragma once
 
+#include <raft/core/bitset.cuh>
+
 #include <cstddef>
 #include <cstdint>
 
-#include <raft/core/bitset.cuh>
-
 namespace raft::neighbors::filtering {
 /**
  * @brief Filter an index with a bitset
diff --git a/cpp/include/raft/neighbors/sample_filter_types.hpp b/cpp/include/raft/neighbors/sample_filter_types.hpp
index 25030f48c8..b0c482d914 100644
--- a/cpp/include/raft/neighbors/sample_filter_types.hpp
+++ b/cpp/include/raft/neighbors/sample_filter_types.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,11 @@
 
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
+
 #include <cstddef>
 #include <cstdint>
 
-#include <raft/core/detail/macros.hpp>
-
 namespace raft::neighbors::filtering {
 
 /* A filter that filters nothing. This is the default behavior. */
diff --git a/cpp/include/raft/neighbors/specializations/detail/ball_cover_lowdim.hpp b/cpp/include/raft/neighbors/specializations/detail/ball_cover_lowdim.hpp
index fa2689e960..fbe38fd261 100644
--- a/cpp/include/raft/neighbors/specializations/detail/ball_cover_lowdim.hpp
+++ b/cpp/include/raft/neighbors/specializations/detail/ball_cover_lowdim.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,11 @@
  * limitations under the License.
  */
 
-#include <cstdint>
 #include <raft/spatial/knn/detail/ball_cover/common.cuh>
 #include <raft/spatial/knn/detail/ball_cover/registers.cuh>
 
+#include <cstdint>
+
 namespace raft {
 namespace spatial {
 namespace knn {
diff --git a/cpp/include/raft/random/detail/make_blobs.cuh b/cpp/include/raft/random/detail/make_blobs.cuh
index 1140b3484f..a9c7fc0b40 100644
--- a/cpp/include/raft/random/detail/make_blobs.cuh
+++ b/cpp/include/raft/random/detail/make_blobs.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,13 +17,16 @@
 #pragma once
 
 #include "permute.cuh"
+
 #include <raft/core/handle.hpp>
 #include <raft/linalg/map.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/random/rng_device.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
+
 #include <vector>
 
 namespace raft {
diff --git a/cpp/include/raft/random/detail/make_regression.cuh b/cpp/include/raft/random/detail/make_regression.cuh
index ecae8832fa..f79534c75a 100644
--- a/cpp/include/raft/random/detail/make_regression.cuh
+++ b/cpp/include/raft/random/detail/make_regression.cuh
@@ -20,8 +20,6 @@
 
 #pragma once
 
-#include <algorithm>
-
 #include <raft/core/resources.hpp>
 #include <raft/linalg/add.cuh>
 #include <raft/linalg/gemm.cuh>
@@ -32,8 +30,11 @@
 #include <raft/random/permute.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <algorithm>
+
 namespace raft::random {
 namespace detail {
 
diff --git a/cpp/include/raft/random/detail/multi_variable_gaussian.cuh b/cpp/include/raft/random/detail/multi_variable_gaussian.cuh
index b88374d5b4..e88cbbdeea 100644
--- a/cpp/include/raft/random/detail/multi_variable_gaussian.cuh
+++ b/cpp/include/raft/random/detail/multi_variable_gaussian.cuh
@@ -16,9 +16,7 @@
 
 #pragma once
 #include "curand_wrappers.hpp"
-#include <cmath>
-#include <memory>
-#include <optional>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cusolver_dn_handle.hpp>
@@ -31,8 +29,14 @@
 #include <raft/random/random_types.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
+
 #include <stdio.h>
+
+#include <cmath>
+#include <memory>
+#include <optional>
 #include <type_traits>
 
 // mvg.cuh takes in matrices that are column major (as in fortran)
diff --git a/cpp/include/raft/random/detail/permute.cuh b/cpp/include/raft/random/detail/permute.cuh
index cf68a746a9..37caa51ad3 100644
--- a/cpp/include/raft/random/detail/permute.cuh
+++ b/cpp/include/raft/random/detail/permute.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,14 @@
 
 #pragma once
 
-#include <cooperative_groups.h>
-#include <memory>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/vectorized.cuh>
 
+#include <cooperative_groups.h>
+
+#include <memory>
+
 namespace raft::random {
 namespace detail {
 
diff --git a/cpp/include/raft/random/detail/rng_device.cuh b/cpp/include/raft/random/detail/rng_device.cuh
index 26bf93c260..12c67679ba 100644
--- a/cpp/include/raft/random/detail/rng_device.cuh
+++ b/cpp/include/raft/random/detail/rng_device.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,10 +20,10 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/integer_utils.hpp>
 
-#include <curand_kernel.h>
-
 #include <rmm/device_uvector.hpp>
 
+#include <curand_kernel.h>
+
 #include <random>
 
 namespace raft {
diff --git a/cpp/include/raft/random/detail/rng_impl_deprecated.cuh b/cpp/include/raft/random/detail/rng_impl_deprecated.cuh
index 8895d22cf0..dd2de15b87 100644
--- a/cpp/include/raft/random/detail/rng_impl_deprecated.cuh
+++ b/cpp/include/raft/random/detail/rng_impl_deprecated.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -22,15 +22,18 @@
 
 #include "rng_device.cuh"
 
-#include <curand_kernel.h>
 #include <raft/core/resources.hpp>
 #include <raft/random/rng_state.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/detail/cub_wrappers.cuh>
 #include <raft/util/scatter.cuh>
-#include <random>
+
 #include <rmm/device_uvector.hpp>
 
+#include <curand_kernel.h>
+
+#include <random>
+
 namespace raft {
 namespace random {
 namespace detail {
diff --git a/cpp/include/raft/random/device/sample.cuh b/cpp/include/raft/random/device/sample.cuh
index 24279f82e4..d0e5200185 100644
--- a/cpp/include/raft/random/device/sample.cuh
+++ b/cpp/include/raft/random/device/sample.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <stdint.h>
-
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/util/cuda_dev_essentials.cuh>
 #include <raft/util/warp_primitives.cuh>
 
+#include <stdint.h>
+
 namespace raft::random::device {
 
 /**
diff --git a/cpp/include/raft/random/make_blobs.cuh b/cpp/include/raft/random/make_blobs.cuh
index 079ab43b74..4fd1f44f64 100644
--- a/cpp/include/raft/random/make_blobs.cuh
+++ b/cpp/include/raft/random/make_blobs.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,11 +20,13 @@
 #pragma once
 
 #include "detail/make_blobs.cuh"
-#include <optional>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
+#include <optional>
+
 namespace raft::random {
 
 /**
diff --git a/cpp/include/raft/random/make_regression.cuh b/cpp/include/raft/random/make_regression.cuh
index 0aa9cc4daa..94d10257f9 100644
--- a/cpp/include/raft/random/make_regression.cuh
+++ b/cpp/include/raft/random/make_regression.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,13 +23,14 @@
 
 #pragma once
 
-#include <algorithm>
-#include <optional>
+#include "detail/make_regression.cuh"
+
 #include <raft/core/mdarray.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 
-#include "detail/make_regression.cuh"
+#include <algorithm>
+#include <optional>
 
 namespace raft::random {
 
diff --git a/cpp/include/raft/random/multi_variable_gaussian.cuh b/cpp/include/raft/random/multi_variable_gaussian.cuh
index eada1c9521..ab3f433422 100644
--- a/cpp/include/raft/random/multi_variable_gaussian.cuh
+++ b/cpp/include/raft/random/multi_variable_gaussian.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,6 +20,7 @@
 #pragma once
 
 #include "detail/multi_variable_gaussian.cuh"
+
 #include <raft/core/resources.hpp>
 #include <raft/random/random_types.hpp>
 
diff --git a/cpp/include/raft/random/permute.cuh b/cpp/include/raft/random/permute.cuh
index d349b68add..7c3a4cc6c8 100644
--- a/cpp/include/raft/random/permute.cuh
+++ b/cpp/include/raft/random/permute.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,10 +21,11 @@
 
 #include "detail/permute.cuh"
 
-#include <optional>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+
+#include <optional>
 #include <type_traits>
 
 namespace raft::random {
diff --git a/cpp/include/raft/random/rmat_rectangular_generator.cuh b/cpp/include/raft/random/rmat_rectangular_generator.cuh
index 90cd9baf81..5598b25c8e 100644
--- a/cpp/include/raft/random/rmat_rectangular_generator.cuh
+++ b/cpp/include/raft/random/rmat_rectangular_generator.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include "detail/rmat_rectangular_generator.cuh"
+
 #include <raft/core/resources.hpp>
 
 namespace raft::random {
diff --git a/cpp/include/raft/random/rng.cuh b/cpp/include/raft/random/rng.cuh
index a946471312..4e63669f98 100644
--- a/cpp/include/raft/random/rng.cuh
+++ b/cpp/include/raft/random/rng.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,11 +19,13 @@
 #include "detail/rng_impl.cuh"
 #include "detail/rng_impl_deprecated.cuh"  // necessary for now (to be removed)
 #include "rng_state.hpp"
-#include <cassert>
-#include <optional>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+
+#include <cassert>
+#include <optional>
 #include <type_traits>
 #include <variant>
 
diff --git a/cpp/include/raft/random/sample_without_replacement.cuh b/cpp/include/raft/random/sample_without_replacement.cuh
index b074f68af6..fad1d4adfa 100644
--- a/cpp/include/raft/random/sample_without_replacement.cuh
+++ b/cpp/include/raft/random/sample_without_replacement.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,11 +18,13 @@
 
 #include "detail/rng_impl.cuh"
 #include "rng_state.hpp"
-#include <cassert>
-#include <optional>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+
+#include <cassert>
+#include <optional>
 #include <type_traits>
 #include <variant>
 
diff --git a/cpp/include/raft/solver/detail/lap_functions.cuh b/cpp/include/raft/solver/detail/lap_functions.cuh
index 30a4961abf..dc39082f63 100644
--- a/cpp/include/raft/solver/detail/lap_functions.cuh
+++ b/cpp/include/raft/solver/detail/lap_functions.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  * Copyright 2020 KETAN DATE & RAKESH NAGI
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
@@ -25,11 +25,11 @@
 #pragma once
 
 #include <raft/core/resource/cuda_stream.hpp>
-#include <raft/solver/linear_assignment_types.hpp>
-
 #include <raft/core/resources.hpp>
 #include <raft/solver/detail/lap_kernels.cuh>
+#include <raft/solver/linear_assignment_types.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/include/raft/solver/linear_assignment.cuh b/cpp/include/raft/solver/linear_assignment.cuh
index a88356a42e..7ee0f5fbc3 100644
--- a/cpp/include/raft/solver/linear_assignment.cuh
+++ b/cpp/include/raft/solver/linear_assignment.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  * Copyright 2020 KETAN DATE & RAKESH NAGI
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
@@ -30,14 +30,14 @@
 
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+#include <raft/solver/detail/lap_functions.cuh>
+#include <raft/solver/linear_assignment_types.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/execution_policy.h>
 #include <thrust/fill.h>
 
-#include <raft/solver/detail/lap_functions.cuh>
-#include <raft/solver/linear_assignment_types.hpp>
-
 namespace raft::solver {
 
 /**
diff --git a/cpp/include/raft/sparse/convert/detail/adj_to_csr.cuh b/cpp/include/raft/sparse/convert/detail/adj_to_csr.cuh
index 4a4453417d..69949a71ed 100644
--- a/cpp/include/raft/sparse/convert/detail/adj_to_csr.cuh
+++ b/cpp/include/raft/sparse/convert/detail/adj_to_csr.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,15 +16,16 @@
 
 #pragma once
 
-#include <cooperative_groups.h>
 #include <raft/core/resource/cuda_stream.hpp>
-
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/device_atomics.cuh>
 #include <raft/util/vectorized.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cooperative_groups.h>
+
 namespace raft {
 namespace sparse {
 namespace convert {
diff --git a/cpp/include/raft/sparse/convert/detail/coo.cuh b/cpp/include/raft/sparse/convert/detail/coo.cuh
index 1302ed1220..0a498bb1ca 100644
--- a/cpp/include/raft/sparse/convert/detail/coo.cuh
+++ b/cpp/include/raft/sparse/convert/detail/coo.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,22 +16,21 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
+#include <raft/sparse/coo.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace convert {
diff --git a/cpp/include/raft/sparse/convert/detail/csr.cuh b/cpp/include/raft/sparse/convert/detail/csr.cuh
index 11e745f680..a5d7de9a07 100644
--- a/cpp/include/raft/sparse/convert/detail/csr.cuh
+++ b/cpp/include/raft/sparse/convert/detail/csr.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,30 +16,29 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cusparse_handle.hpp>
-
 #include <raft/core/resources.hpp>
+#include <raft/sparse/coo.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
+#include <raft/sparse/linalg/degree.cuh>
+#include <raft/sparse/op/row_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
-#include <raft/sparse/linalg/degree.cuh>
-#include <raft/sparse/op/row_op.cuh>
-
 namespace raft {
 namespace sparse {
 namespace convert {
diff --git a/cpp/include/raft/sparse/convert/detail/dense.cuh b/cpp/include/raft/sparse/convert/detail/dense.cuh
index f3db8a36c1..e60e494d34 100644
--- a/cpp/include/raft/sparse/convert/detail/dense.cuh
+++ b/cpp/include/raft/sparse/convert/detail/dense.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,23 +16,23 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <rmm/device_uvector.hpp>
+
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-#include <rmm/device_uvector.hpp>
-
 namespace raft {
 namespace sparse {
 namespace convert {
diff --git a/cpp/include/raft/sparse/detail/coo.cuh b/cpp/include/raft/sparse/detail/coo.cuh
index cbcbee0139..91ba363168 100644
--- a/cpp/include/raft/sparse/detail/coo.cuh
+++ b/cpp/include/raft/sparse/detail/coo.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,12 @@
  * limitations under the License.
  */
 
-#include <iostream>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <iostream>
+
 #pragma once
 
 namespace raft {
diff --git a/cpp/include/raft/sparse/detail/csr.cuh b/cpp/include/raft/sparse/detail/csr.cuh
index 4f8496a026..9901e89b8c 100644
--- a/cpp/include/raft/sparse/detail/csr.cuh
+++ b/cpp/include/raft/sparse/detail/csr.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,24 +16,24 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace detail {
diff --git a/cpp/include/raft/sparse/detail/cusparse_wrappers.h b/cpp/include/raft/sparse/detail/cusparse_wrappers.h
index cc3ae3ab87..08efbb7106 100644
--- a/cpp/include/raft/sparse/detail/cusparse_wrappers.h
+++ b/cpp/include/raft/sparse/detail/cusparse_wrappers.h
@@ -16,12 +16,14 @@
 
 #pragma once
 
-#include <cusparse.h>
 #include <raft/core/cusparse_macros.hpp>
 #include <raft/core/error.hpp>
 #include <raft/linalg/transpose.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cusparse.h>
+
 namespace raft {
 namespace sparse {
 namespace detail {
diff --git a/cpp/include/raft/sparse/distance/detail/bin_distance.cuh b/cpp/include/raft/sparse/distance/detail/bin_distance.cuh
index ae16b1adcc..7a2396c2cd 100644
--- a/cpp/include/raft/sparse/distance/detail/bin_distance.cuh
+++ b/cpp/include/raft/sparse/distance/detail/bin_distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,17 +16,19 @@
 
 #pragma once
 
-#include <limits.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "common.hpp"
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/sparse/detail/utils.h>
 #include <raft/sparse/distance/detail/ip_distance.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <limits.h>
+
 #include <nvfunctional>
 
 namespace raft {
diff --git a/cpp/include/raft/sparse/distance/detail/coo_spmv.cuh b/cpp/include/raft/sparse/distance/detail/coo_spmv.cuh
index c0d5fbc365..b0469f559a 100644
--- a/cpp/include/raft/sparse/distance/detail/coo_spmv.cuh
+++ b/cpp/include/raft/sparse/distance/detail/coo_spmv.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,24 +16,22 @@
 
 #pragma once
 
+#include "../../csr.hpp"
+#include "../../detail/utils.h"
+#include "common.hpp"
 #include "coo_spmv_strategies/dense_smem_strategy.cuh"
 #include "coo_spmv_strategies/hash_strategy.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include "../../csr.hpp"
-#include "../../detail/utils.h"
-#include "common.hpp"
-
+#include <cusparse_v2.h>
 #include <limits.h>
 
 #include <nvfunctional>
 
-#include <cusparse_v2.h>
-
 namespace raft {
 namespace sparse {
 namespace distance {
diff --git a/cpp/include/raft/sparse/distance/detail/coo_spmv_kernel.cuh b/cpp/include/raft/sparse/distance/detail/coo_spmv_kernel.cuh
index 43e8dea3ee..d8f5b2eca5 100644
--- a/cpp/include/raft/sparse/distance/detail/coo_spmv_kernel.cuh
+++ b/cpp/include/raft/sparse/distance/detail/coo_spmv_kernel.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,10 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-
 #include <cub/block/block_load.cuh>
 #include <cub/block/block_radix_sort.cuh>
 #include <cub/block/block_store.cuh>
+#include <cub/cub.cuh>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/base_strategy.cuh b/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/base_strategy.cuh
index 1c2f83c69b..fc5881e2d4 100644
--- a/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/base_strategy.cuh
+++ b/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/base_strategy.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,6 +20,7 @@
 #include "../coo_spmv_kernel.cuh"
 #include "../utils.cuh"
 #include "coo_mask_row_iterators.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 
 #include <rmm/device_uvector.hpp>
diff --git a/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/hash_strategy.cuh b/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/hash_strategy.cuh
index 20c722f3e0..e271f2cdbe 100644
--- a/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/hash_strategy.cuh
+++ b/cpp/include/raft/sparse/distance/detail/coo_spmv_strategies/hash_strategy.cuh
@@ -17,11 +17,11 @@
 #pragma once
 
 #include "base_strategy.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
 
 #include <cuco/static_map.cuh>
-
 #include <thrust/copy.h>
 #include <thrust/iterator/counting_iterator.h>
 
diff --git a/cpp/include/raft/sparse/distance/detail/ip_distance.cuh b/cpp/include/raft/sparse/distance/detail/ip_distance.cuh
index 39e67acdea..84229120ea 100644
--- a/cpp/include/raft/sparse/distance/detail/ip_distance.cuh
+++ b/cpp/include/raft/sparse/distance/detail/ip_distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,22 +16,24 @@
 
 #pragma once
 
-#include <limits.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/distance/distance_types.hpp>
-#include <raft/sparse/detail/cusparse_wrappers.h>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-
 #include "common.hpp"
+
 #include <raft/core/operators.cuh>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/distance/distance_types.hpp>
 #include <raft/sparse/convert/coo.cuh>
+#include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/sparse/detail/utils.h>
 #include <raft/sparse/distance/detail/coo_spmv.cuh>
 #include <raft/sparse/linalg/transpose.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <limits.h>
+
 #include <nvfunctional>
 
 namespace raft {
diff --git a/cpp/include/raft/sparse/distance/detail/l2_distance.cuh b/cpp/include/raft/sparse/distance/detail/l2_distance.cuh
index c198f91a31..93d1f048ec 100644
--- a/cpp/include/raft/sparse/distance/detail/l2_distance.cuh
+++ b/cpp/include/raft/sparse/distance/detail/l2_distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,26 +16,26 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/spatial/knn/knn.cuh>
-
 #include "common.hpp"
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/sparse/detail/utils.h>
 #include <raft/sparse/distance/detail/ip_distance.cuh>
+#include <raft/spatial/knn/knn.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
-#include <rmm/device_uvector.hpp>
 
-#include <nvfunctional>
+#include <rmm/device_uvector.hpp>
 
 #include <thrust/for_each.h>
 #include <thrust/iterator/counting_iterator.h>
 
 #include <algorithm>
+#include <nvfunctional>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/include/raft/sparse/distance/detail/lp_distance.cuh b/cpp/include/raft/sparse/distance/detail/lp_distance.cuh
index ff9534a157..b178e02c34 100644
--- a/cpp/include/raft/sparse/distance/detail/lp_distance.cuh
+++ b/cpp/include/raft/sparse/distance/detail/lp_distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,25 +16,24 @@
 
 #pragma once
 
-#include <limits.h>
-#include <raft/core/resource/cuda_stream.hpp>
+#include "common.hpp"
 
 #include <raft/core/operators.cuh>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-#include <rmm/device_uvector.hpp>
-
+#include <raft/sparse/convert/coo.cuh>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/detail/utils.h>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
 
-#include "common.hpp"
-#include <raft/sparse/convert/coo.cuh>
+#include <rmm/device_uvector.hpp>
 
-#include <nvfunctional>
+#include <limits.h>
 
 #include <algorithm>
+#include <nvfunctional>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/include/raft/sparse/distance/distance.cuh b/cpp/include/raft/sparse/distance/distance.cuh
index 702846f586..ead44f0c51 100644
--- a/cpp/include/raft/sparse/distance/distance.cuh
+++ b/cpp/include/raft/sparse/distance/distance.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,17 +20,16 @@
 #pragma once
 
 #include "detail/common.hpp"
-#include <unordered_set>
 
 #include <raft/core/device_csr_matrix.hpp>
-
 #include <raft/distance/distance_types.hpp>
-
 #include <raft/sparse/distance/detail/bin_distance.cuh>
 #include <raft/sparse/distance/detail/ip_distance.cuh>
 #include <raft/sparse/distance/detail/l2_distance.cuh>
 #include <raft/sparse/distance/detail/lp_distance.cuh>
 
+#include <unordered_set>
+
 namespace raft {
 namespace sparse {
 namespace distance {
diff --git a/cpp/include/raft/sparse/linalg/detail/add.cuh b/cpp/include/raft/sparse/linalg/detail/add.cuh
index 72dd4a60f9..4be15693ef 100644
--- a/cpp/include/raft/sparse/linalg/detail/add.cuh
+++ b/cpp/include/raft/sparse/linalg/detail/add.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,25 +16,24 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
-
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 #include <rmm/exec_policy.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace linalg {
diff --git a/cpp/include/raft/sparse/linalg/detail/cusparse_utils.hpp b/cpp/include/raft/sparse/linalg/detail/cusparse_utils.hpp
index b15614905b..c10c0de426 100644
--- a/cpp/include/raft/sparse/linalg/detail/cusparse_utils.hpp
+++ b/cpp/include/raft/sparse/linalg/detail/cusparse_utils.hpp
@@ -16,14 +16,14 @@
 
 #pragma once
 
-#include <type_traits>
-
 #include <raft/core/device_csr_matrix.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/linalg/linalg_types.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/util/input_validation.hpp>
 
+#include <type_traits>
+
 namespace raft {
 namespace sparse {
 namespace linalg {
diff --git a/cpp/include/raft/sparse/linalg/detail/degree.cuh b/cpp/include/raft/sparse/linalg/detail/degree.cuh
index a68968ea97..df31192cf7 100644
--- a/cpp/include/raft/sparse/linalg/detail/degree.cuh
+++ b/cpp/include/raft/sparse/linalg/detail/degree.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,14 @@
 
 #pragma once
 
+#include <raft/sparse/coo.hpp>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/device_atomics.cuh>
 
 #include <cuda_runtime.h>
-#include <stdio.h>
 
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
+#include <stdio.h>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/include/raft/sparse/linalg/detail/norm.cuh b/cpp/include/raft/sparse/linalg/detail/norm.cuh
index d68e1e992f..3702111f83 100644
--- a/cpp/include/raft/sparse/linalg/detail/norm.cuh
+++ b/cpp/include/raft/sparse/linalg/detail/norm.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,27 +16,25 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/common/nvtx.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/linalg/norm_types.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
+#include <raft/sparse/op/row_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <raft/sparse/op/row_op.cuh>
-
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <iostream>
 #include <limits>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace linalg {
diff --git a/cpp/include/raft/sparse/linalg/detail/spectral.cuh b/cpp/include/raft/sparse/linalg/detail/spectral.cuh
index 51836ca9aa..a1642d1455 100644
--- a/cpp/include/raft/sparse/linalg/detail/spectral.cuh
+++ b/cpp/include/raft/sparse/linalg/detail/spectral.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,15 @@
  */
 
 #include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cudart_utils.hpp>
-
+#include <raft/sparse/convert/csr.cuh>
+#include <raft/sparse/coo.hpp>
 #include <raft/spectral/cluster_solvers.cuh>
 #include <raft/spectral/eigen_solvers.cuh>
 #include <raft/spectral/partition.cuh>
 #include <raft/util/cuda_utils.cuh>
-#include <rmm/device_uvector.hpp>
+#include <raft/util/cudart_utils.hpp>
 
-#include <raft/sparse/convert/csr.cuh>
-#include <raft/sparse/coo.hpp>
+#include <rmm/device_uvector.hpp>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/include/raft/sparse/linalg/detail/symmetrize.cuh b/cpp/include/raft/sparse/linalg/detail/symmetrize.cuh
index 6febf0aadc..d343bcbf66 100644
--- a/cpp/include/raft/sparse/linalg/detail/symmetrize.cuh
+++ b/cpp/include/raft/sparse/linalg/detail/symmetrize.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,31 +16,30 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/core/resource/cuda_stream.hpp>
-
+#include <raft/sparse/convert/csr.cuh>
+#include <raft/sparse/coo.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
+#include <raft/sparse/op/reduce.cuh>
+#include <raft/sparse/op/sort.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+#include <raft/util/device_atomics.cuh>
+
 #include <rmm/device_uvector.hpp>
 #include <rmm/exec_policy.hpp>
 
-#include <raft/sparse/op/sort.cuh>
-#include <raft/util/device_atomics.cuh>
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/convert/csr.cuh>
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
-#include <raft/sparse/op/reduce.cuh>
-
 namespace raft {
 namespace sparse {
 namespace linalg {
diff --git a/cpp/include/raft/sparse/linalg/detail/transpose.h b/cpp/include/raft/sparse/linalg/detail/transpose.h
index 1484804348..3a646b9a6e 100644
--- a/cpp/include/raft/sparse/linalg/detail/transpose.h
+++ b/cpp/include/raft/sparse/linalg/detail/transpose.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,24 +16,23 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
-
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace linalg {
diff --git a/cpp/include/raft/sparse/neighbors/detail/cross_component_nn.cuh b/cpp/include/raft/sparse/neighbors/detail/cross_component_nn.cuh
index 749152ee07..a47d5a6f34 100644
--- a/cpp/include/raft/sparse/neighbors/detail/cross_component_nn.cuh
+++ b/cpp/include/raft/sparse/neighbors/detail/cross_component_nn.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,6 @@
  */
 #pragma once
 
-#include <cstdint>
-#include <cub/cub.cuh>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/kvp.hpp>
@@ -33,25 +31,24 @@
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/linalg/symmetrize.cuh>
 #include <raft/sparse/op/reduce.cuh>
-
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/fast_int_div.cuh>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
 #include <thrust/copy.h>
 #include <thrust/device_ptr.h>
+#include <thrust/gather.h>
 #include <thrust/iterator/counting_iterator.h>
 #include <thrust/iterator/zip_iterator.h>
 #include <thrust/scan.h>
+#include <thrust/scatter.h>
 #include <thrust/sort.h>
 #include <thrust/transform.h>
 #include <thrust/tuple.h>
 
-#include <thrust/gather.h>
-#include <thrust/scatter.h>
-
-#include <cub/cub.cuh>
-
+#include <cstdint>
 #include <limits>
 
 namespace raft::sparse::neighbors::detail {
diff --git a/cpp/include/raft/sparse/neighbors/detail/knn.cuh b/cpp/include/raft/sparse/neighbors/detail/knn.cuh
index f9d019931a..68bba31360 100644
--- a/cpp/include/raft/sparse/neighbors/detail/knn.cuh
+++ b/cpp/include/raft/sparse/neighbors/detail/knn.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,19 +17,18 @@
 #pragma once
 
 #include <raft/core/resource/cuda_stream.hpp>
-#include <rmm/device_uvector.hpp>
-
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/unary_op.cuh>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-
 #include <raft/matrix/select_k.cuh>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/detail/utils.h>
 #include <raft/sparse/distance/distance.cuh>
 #include <raft/sparse/op/slice.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+
+#include <rmm/device_uvector.hpp>
 
 #include <algorithm>
 
diff --git a/cpp/include/raft/sparse/neighbors/detail/knn_graph.cuh b/cpp/include/raft/sparse/neighbors/detail/knn_graph.cuh
index 4344025065..4e46904c83 100644
--- a/cpp/include/raft/sparse/neighbors/detail/knn_graph.cuh
+++ b/cpp/include/raft/sparse/neighbors/detail/knn_graph.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,15 +17,13 @@
 #pragma once
 
 #include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-
+#include <raft/distance/distance_types.hpp>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/linalg/symmetrize.cuh>
-
 #include <raft/spatial/knn/knn.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
 
-#include <raft/distance/distance_types.hpp>
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/device_ptr.h>
diff --git a/cpp/include/raft/sparse/op/detail/filter.cuh b/cpp/include/raft/sparse/op/detail/filter.cuh
index aa375443f5..3df85e6871 100644
--- a/cpp/include/raft/sparse/op/detail/filter.cuh
+++ b/cpp/include/raft/sparse/op/detail/filter.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,28 +16,27 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
-
+#include <raft/sparse/coo.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
+#include <raft/sparse/linalg/degree.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 #include <rmm/exec_policy.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/reduce.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 
 #include <algorithm>
 #include <cstdio>
 #include <iostream>
 
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
-#include <raft/sparse/linalg/degree.cuh>
-
 namespace raft {
 namespace sparse {
 namespace op {
diff --git a/cpp/include/raft/sparse/op/detail/reduce.cuh b/cpp/include/raft/sparse/op/detail/reduce.cuh
index 8a1a63e424..1e5dd87958 100644
--- a/cpp/include/raft/sparse/op/detail/reduce.cuh
+++ b/cpp/include/raft/sparse/op/detail/reduce.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,30 +16,29 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
-
+#include <raft/sparse/convert/csr.cuh>
+#include <raft/sparse/coo.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
+#include <raft/sparse/op/sort.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
-
-#include <raft/sparse/op/sort.cuh>
 #include <raft/util/device_atomics.cuh>
+
+#include <rmm/device_uvector.hpp>
+
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
-#include <rmm/device_uvector.hpp>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/convert/csr.cuh>
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace op {
diff --git a/cpp/include/raft/sparse/op/detail/row_op.cuh b/cpp/include/raft/sparse/op/detail/row_op.cuh
index 256ae28f8c..73dbec6389 100644
--- a/cpp/include/raft/sparse/op/detail/row_op.cuh
+++ b/cpp/include/raft/sparse/op/detail/row_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,23 +16,21 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
-
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace op {
diff --git a/cpp/include/raft/sparse/op/detail/slice.cuh b/cpp/include/raft/sparse/op/detail/slice.cuh
index 4d2f1a4195..c2d786c9fe 100644
--- a/cpp/include/raft/sparse/op/detail/slice.cuh
+++ b/cpp/include/raft/sparse/op/detail/slice.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,25 +16,23 @@
 
 #pragma once
 
-#include <cusparse_v2.h>
-
 #include <raft/core/operators.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/scan.h>
 
-#include <cuda_runtime.h>
+#include <cusparse_v2.h>
 #include <stdio.h>
 
 #include <algorithm>
 #include <iostream>
 
-#include <raft/sparse/detail/utils.h>
-
 namespace raft {
 namespace sparse {
 namespace op {
diff --git a/cpp/include/raft/sparse/op/detail/sort.h b/cpp/include/raft/sparse/op/detail/sort.h
index 2f73671132..85ae825035 100644
--- a/cpp/include/raft/sparse/op/detail/sort.h
+++ b/cpp/include/raft/sparse/op/detail/sort.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,8 +21,10 @@
 #include <raft/sparse/detail/utils.h>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/exec_policy.hpp>
 
+#include <cuda_runtime.h>
 #include <thrust/device_ptr.h>
 #include <thrust/iterator/zip_iterator.h>
 #include <thrust/scan.h>
@@ -31,8 +33,6 @@
 
 #include <cusparse_v2.h>
 
-#include <cuda_runtime.h>
-
 #include <algorithm>
 
 namespace raft {
diff --git a/cpp/include/raft/sparse/solver/detail/lanczos.cuh b/cpp/include/raft/sparse/solver/detail/lanczos.cuh
index 74f8931a17..9ecb4b729f 100644
--- a/cpp/include/raft/sparse/solver/detail/lanczos.cuh
+++ b/cpp/include/raft/sparse/solver/detail/lanczos.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,14 +19,8 @@
 // for cmath:
 #define _USE_MATH_DEFINES
 
-#include <cmath>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
-
-#include <cuda.h>
-#include <curand.h>
-
 #include <raft/core/resources.hpp>
 #include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/spectral/detail/lapack.hpp>
@@ -34,6 +28,13 @@
 #include <raft/spectral/matrix_wrappers.hpp>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cuda.h>
+
+#include <curand.h>
+
+#include <cmath>
+#include <vector>
+
 namespace raft::sparse::solver::detail {
 
 // curandGeneratorNormalX
diff --git a/cpp/include/raft/sparse/solver/detail/mst_kernels.cuh b/cpp/include/raft/sparse/solver/detail/mst_kernels.cuh
index 5b56ea45f0..968eee4002 100644
--- a/cpp/include/raft/sparse/solver/detail/mst_kernels.cuh
+++ b/cpp/include/raft/sparse/solver/detail/mst_kernels.cuh
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,11 +18,10 @@
 #pragma once
 
 #include <raft/sparse/solver/detail/mst_utils.cuh>
+#include <raft/util/device_atomics.cuh>
 
 #include <limits>
 
-#include <raft/util/device_atomics.cuh>
-
 namespace raft::sparse::solver::detail {
 
 template <typename vertex_t, typename edge_t, typename alteration_t>
diff --git a/cpp/include/raft/sparse/solver/detail/mst_solver_inl.cuh b/cpp/include/raft/sparse/solver/detail/mst_solver_inl.cuh
index 0182668117..7ce8871f06 100644
--- a/cpp/include/raft/sparse/solver/detail/mst_solver_inl.cuh
+++ b/cpp/include/raft/sparse/solver/detail/mst_solver_inl.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,10 @@
 
 #pragma once
 
-#include <curand.h>
 #include <raft/core/resource/device_properties.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
-
 #include <raft/sparse/solver/detail/mst_kernels.cuh>
 #include <raft/sparse/solver/detail/mst_utils.cuh>
-
 #include <raft/util/cudart_utils.hpp>
 
 #include <rmm/device_scalar.hpp>
@@ -43,6 +40,8 @@
 #include <thrust/tuple.h>
 #include <thrust/unique.h>
 
+#include <curand.h>
+
 #include <iostream>
 
 namespace raft::sparse::solver {
diff --git a/cpp/include/raft/sparse/solver/detail/mst_utils.cuh b/cpp/include/raft/sparse/solver/detail/mst_utils.cuh
index a33141192b..f375623791 100644
--- a/cpp/include/raft/sparse/solver/detail/mst_utils.cuh
+++ b/cpp/include/raft/sparse/solver/detail/mst_utils.cuh
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,9 +17,10 @@
 
 #pragma once
 
-#include <iostream>
 #include <rmm/device_uvector.hpp>
 
+#include <iostream>
+
 namespace raft::sparse::solver::detail {
 
 template <typename idx_t>
diff --git a/cpp/include/raft/sparse/solver/mst_solver.cuh b/cpp/include/raft/sparse/solver/mst_solver.cuh
index 193431251f..8317e66d4a 100644
--- a/cpp/include/raft/sparse/solver/mst_solver.cuh
+++ b/cpp/include/raft/sparse/solver/mst_solver.cuh
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,6 +18,7 @@
 #pragma once
 
 #include <raft/core/resources.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/include/raft/spatial/knn/ann_common.h b/cpp/include/raft/spatial/knn/ann_common.h
index 73a70cbefe..722c01d561 100644
--- a/cpp/include/raft/spatial/knn/ann_common.h
+++ b/cpp/include/raft/spatial/knn/ann_common.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,9 +23,9 @@
 
 #include "detail/processing.hpp"
 #include "ivf_flat_types.hpp"
-#include <raft/neighbors/ivf_pq_types.hpp>
 
 #include <raft/distance/distance_types.hpp>
+#include <raft/neighbors/ivf_pq_types.hpp>
 
 namespace raft {
 namespace spatial {
diff --git a/cpp/include/raft/spatial/knn/detail/ann_quantized.cuh b/cpp/include/raft/spatial/knn/detail/ann_quantized.cuh
index 964292f6cb..041ab225f9 100644
--- a/cpp/include/raft/spatial/knn/detail/ann_quantized.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ann_quantized.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,19 +18,17 @@
 
 #include "../ann_common.h"
 #include "../ivf_flat.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "processing.cuh"
-#include <raft/core/operators.hpp>
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
 
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 #include <raft/label/classlabels.cuh>
 #include <raft/neighbors/ivf_pq.cuh>
-
-#include <raft/core/device_mdspan.hpp>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
 
 #include <rmm/cuda_stream_view.hpp>
 
diff --git a/cpp/include/raft/spatial/knn/detail/ann_utils.cuh b/cpp/include/raft/spatial/knn/detail/ann_utils.cuh
index d75c7bf531..d862e586e3 100644
--- a/cpp/include/raft/spatial/knn/detail/ann_utils.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ann_utils.cuh
@@ -26,11 +26,11 @@
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_fp16.hpp>
+
 #include <memory>
 #include <optional>
 
-#include <cuda_fp16.hpp>
-
 namespace raft::spatial::knn::detail::utils {
 
 /** Whether pointers are accessible on the device or on the host. */
diff --git a/cpp/include/raft/spatial/knn/detail/ball_cover.cuh b/cpp/include/raft/spatial/knn/detail/ball_cover.cuh
index 879f54fd81..5c91b1ad6d 100644
--- a/cpp/include/raft/spatial/knn/detail/ball_cover.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ball_cover.cuh
@@ -16,27 +16,21 @@
 
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-#include <raft/core/resources.hpp>
-
 #include "../ball_cover_types.hpp"
 #include "ball_cover/common.cuh"
 #include "ball_cover/registers.cuh"
 #include "haversine_distance.cuh"
 
-#include <cstdint>
-#include <limits.h>
-
-#include <raft/util/cuda_utils.cuh>
-
-#include <raft/neighbors/detail/faiss_select/key_value_block_select.cuh>
-
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
+#include <raft/core/resources.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/matrix/copy.cuh>
 #include <raft/neighbors/brute_force.cuh>
+#include <raft/neighbors/detail/faiss_select/key_value_block_select.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/sparse/convert/csr.cuh>
+#include <raft/util/cuda_utils.cuh>
 
 #include <rmm/device_uvector.hpp>
 #include <rmm/exec_policy.hpp>
@@ -51,6 +45,10 @@
 #include <thrust/sort.h>
 #include <thrust/tuple.h>
 
+#include <limits.h>
+
+#include <cstdint>
+
 namespace raft {
 namespace spatial {
 namespace knn {
diff --git a/cpp/include/raft/spatial/knn/detail/ball_cover/common.cuh b/cpp/include/raft/spatial/knn/detail/ball_cover/common.cuh
index ce72b2648f..00c34b851d 100644
--- a/cpp/include/raft/spatial/knn/detail/ball_cover/common.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ball_cover/common.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,10 +18,12 @@
 
 #include "../haversine_distance.cuh"
 #include "registers_types.cuh"
-#include <cstdint>
+
 #include <thrust/functional.h>
 #include <thrust/tuple.h>
 
+#include <cstdint>
+
 namespace raft {
 namespace spatial {
 namespace knn {
diff --git a/cpp/include/raft/spatial/knn/detail/ball_cover/registers-ext.cuh b/cpp/include/raft/spatial/knn/detail/ball_cover/registers-ext.cuh
index 2ed6ee3284..81500a0eae 100644
--- a/cpp/include/raft/spatial/knn/detail/ball_cover/registers-ext.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ball_cover/registers-ext.cuh
@@ -16,11 +16,13 @@
 
 #pragma once
 
-#include "../../ball_cover_types.hpp"   // BallCoverIndex
-#include "registers_types.cuh"          // DistFunc
-#include <cstdint>                      // uint32_t
+#include "../../ball_cover_types.hpp"  // BallCoverIndex
+#include "registers_types.cuh"         // DistFunc
+
 #include <raft/util/raft_explicit.hpp>  //RAFT_EXPLICIT
 
+#include <cstdint>  // uint32_t
+
 #if defined(RAFT_EXPLICIT_INSTANTIATE_ONLY)
 
 namespace raft::spatial::knn::detail {
diff --git a/cpp/include/raft/spatial/knn/detail/ball_cover/registers-inl.cuh b/cpp/include/raft/spatial/knn/detail/ball_cover/registers-inl.cuh
index 8b4e8f287e..e4acd1dc4c 100644
--- a/cpp/include/raft/spatial/knn/detail/ball_cover/registers-inl.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ball_cover/registers-inl.cuh
@@ -16,17 +16,13 @@
 
 #pragma once
 
-#include "common.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/thrust_policy.hpp>
-
 #include "../../ball_cover_types.hpp"
 #include "../haversine_distance.cuh"
+#include "common.cuh"
 #include "registers_types.cuh"  // DistFunc
 
-#include <cstdint>
-#include <limits.h>
-
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/neighbors/detail/faiss_select/key_value_block_select.cuh>
 #include <raft/util/cuda_utils.cuh>
 
@@ -34,6 +30,10 @@
 #include <thrust/fill.h>
 #include <thrust/scan.h>
 
+#include <limits.h>
+
+#include <cstdint>
+
 namespace raft {
 namespace spatial {
 namespace knn {
diff --git a/cpp/include/raft/spatial/knn/detail/ball_cover/registers_types.cuh b/cpp/include/raft/spatial/knn/detail/ball_cover/registers_types.cuh
index 7f4268d2dc..f38a3eeec9 100644
--- a/cpp/include/raft/spatial/knn/detail/ball_cover/registers_types.cuh
+++ b/cpp/include/raft/spatial/knn/detail/ball_cover/registers_types.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,7 +17,8 @@
 #pragma once
 
 #include "../haversine_distance.cuh"  // compute_haversine
-#include <cstdint>                    // uint32_t
+
+#include <cstdint>  // uint32_t
 
 namespace raft {
 namespace spatial {
diff --git a/cpp/include/raft/spatial/knn/detail/fused_l2_knn-ext.cuh b/cpp/include/raft/spatial/knn/detail/fused_l2_knn-ext.cuh
index 0eca119450..1968bdb205 100644
--- a/cpp/include/raft/spatial/knn/detail/fused_l2_knn-ext.cuh
+++ b/cpp/include/raft/spatial/knn/detail/fused_l2_knn-ext.cuh
@@ -15,11 +15,12 @@
  */
 #pragma once
 
-#include <cstddef>                           // size_t
-#include <cstdint>                           // uint32_t
 #include <raft/distance/distance_types.hpp>  // DistanceType
 #include <raft/util/raft_explicit.hpp>       // RAFT_EXPLICIT
 
+#include <cstddef>  // size_t
+#include <cstdint>  // uint32_t
+
 #if defined(RAFT_EXPLICIT_INSTANTIATE_ONLY)
 
 namespace raft::spatial::knn::detail {
diff --git a/cpp/include/raft/spatial/knn/detail/fused_l2_knn-inl.cuh b/cpp/include/raft/spatial/knn/detail/fused_l2_knn-inl.cuh
index 0c9f0059f9..ab64614e68 100644
--- a/cpp/include/raft/spatial/knn/detail/fused_l2_knn-inl.cuh
+++ b/cpp/include/raft/spatial/knn/detail/fused_l2_knn-inl.cuh
@@ -14,12 +14,15 @@
  * limitations under the License.
  */
 #pragma once
-#include <cub/cub.cuh>
-#include <limits>
 #include <raft/linalg/norm.cuh>
 #include <raft/neighbors/detail/faiss_select/Select.cuh>
+
+#include <cub/cub.cuh>
+
+#include <limits>
 // TODO: Need to hide the PairwiseDistance class impl and expose to public API
 #include "processing.cuh"
+
 #include <raft/core/operators.hpp>
 #include <raft/distance/detail/distance.cuh>
 #include <raft/distance/detail/distance_ops/l2_exp.cuh>
diff --git a/cpp/include/raft/spatial/knn/detail/haversine_distance.cuh b/cpp/include/raft/spatial/knn/detail/haversine_distance.cuh
index 5fb912843d..98d0919906 100644
--- a/cpp/include/raft/spatial/knn/detail/haversine_distance.cuh
+++ b/cpp/include/raft/spatial/knn/detail/haversine_distance.cuh
@@ -16,13 +16,12 @@
 
 #pragma once
 
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-#include <raft/util/pow2_utils.cuh>
-
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/detail/faiss_select/Select.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+#include <raft/util/pow2_utils.cuh>
 
 namespace raft {
 namespace spatial {
diff --git a/cpp/include/raft/spatial/knn/detail/processing.cuh b/cpp/include/raft/spatial/knn/detail/processing.cuh
index b4b1cb2c14..072e4ebd90 100644
--- a/cpp/include/raft/spatial/knn/detail/processing.cuh
+++ b/cpp/include/raft/spatial/knn/detail/processing.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,6 +24,7 @@
 #include <raft/linalg/unary_op.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/stats/mean_center.cuh>
+
 #include <rmm/device_uvector.hpp>
 
 namespace raft {
diff --git a/cpp/include/raft/spectral/cluster_solvers.cuh b/cpp/include/raft/spectral/cluster_solvers.cuh
index ecd14ebf42..b693ac4af3 100644
--- a/cpp/include/raft/spectral/cluster_solvers.cuh
+++ b/cpp/include/raft/spectral/cluster_solvers.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,6 +21,7 @@
 
 #include <raft/cluster/kmeans.cuh>
 #include <raft/core/resource/thrust_policy.hpp>
+
 #include <utility>  // for std::pair
 
 namespace raft {
diff --git a/cpp/include/raft/spectral/cluster_solvers_deprecated.cuh b/cpp/include/raft/spectral/cluster_solvers_deprecated.cuh
index 008f1e5d13..40b0324548 100644
--- a/cpp/include/raft/spectral/cluster_solvers_deprecated.cuh
+++ b/cpp/include/raft/spectral/cluster_solvers_deprecated.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -25,6 +25,7 @@
 #pragma once
 
 #include <raft/cluster/kmeans_deprecated.cuh>
+
 #include <utility>  // for std::pair
 
 namespace raft {
diff --git a/cpp/include/raft/spectral/detail/lapack.hpp b/cpp/include/raft/spectral/detail/lapack.hpp
index 2d9e5ae9a4..203e6e50f7 100644
--- a/cpp/include/raft/spectral/detail/lapack.hpp
+++ b/cpp/include/raft/spectral/detail/lapack.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,12 @@
  */
 
 #pragma once
-#include <cusolverDn.h>
-
 #include <raft/core/error.hpp>
 #include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/linalg/detail/cusolver_wrappers.hpp>
 
+#include <cusolverDn.h>
+
 // for now; TODO: check if/where this `define` should be;
 //
 #define USE_LAPACK
diff --git a/cpp/include/raft/spectral/detail/matrix_wrappers.hpp b/cpp/include/raft/spectral/detail/matrix_wrappers.hpp
index 6de0dced8a..30dd6e5e69 100644
--- a/cpp/include/raft/spectral/detail/matrix_wrappers.hpp
+++ b/cpp/include/raft/spectral/detail/matrix_wrappers.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,15 +23,15 @@
 #include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cuda/functional>
 #include <thrust/execution_policy.h>
 #include <thrust/fill.h>
 #include <thrust/reduce.h>
 #include <thrust/system/cuda/execution_policy.h>
 
-#include <cuda/functional>
-
 #include <algorithm>
 
 // =========================================================
diff --git a/cpp/include/raft/spectral/detail/modularity_maximization.hpp b/cpp/include/raft/spectral/detail/modularity_maximization.hpp
index f60fbf863e..2a3b5cf36c 100644
--- a/cpp/include/raft/spectral/detail/modularity_maximization.hpp
+++ b/cpp/include/raft/spectral/detail/modularity_maximization.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,23 +16,23 @@
 
 #pragma once
 
-#include <math.h>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
-#include <stdio.h>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
+#include <raft/spectral/cluster_solvers.cuh>
+#include <raft/spectral/detail/spectral_util.cuh>
+#include <raft/spectral/eigen_solvers.cuh>
+#include <raft/spectral/matrix_wrappers.hpp>
 
 #include <cuda.h>
 #include <thrust/fill.h>
 #include <thrust/reduce.h>
 #include <thrust/transform.h>
 
-#include <tuple>
+#include <math.h>
+#include <stdio.h>
 
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-#include <raft/spectral/cluster_solvers.cuh>
-#include <raft/spectral/detail/spectral_util.cuh>
-#include <raft/spectral/eigen_solvers.cuh>
-#include <raft/spectral/matrix_wrappers.hpp>
+#include <tuple>
 
 namespace raft {
 namespace spectral {
diff --git a/cpp/include/raft/spectral/detail/partition.hpp b/cpp/include/raft/spectral/detail/partition.hpp
index ed59ad3149..f5fc40aad6 100644
--- a/cpp/include/raft/spectral/detail/partition.hpp
+++ b/cpp/include/raft/spectral/detail/partition.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,23 +15,23 @@
  */
 #pragma once
 
-#include <math.h>
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
-#include <stdio.h>
+#include <raft/linalg/detail/cublas_wrappers.hpp>
+#include <raft/spectral/cluster_solvers.cuh>
+#include <raft/spectral/detail/spectral_util.cuh>
+#include <raft/spectral/eigen_solvers.cuh>
+#include <raft/spectral/matrix_wrappers.hpp>
 
 #include <cuda.h>
 #include <thrust/fill.h>
 #include <thrust/reduce.h>
 #include <thrust/transform.h>
 
-#include <tuple>
+#include <math.h>
+#include <stdio.h>
 
-#include <raft/linalg/detail/cublas_wrappers.hpp>
-#include <raft/spectral/cluster_solvers.cuh>
-#include <raft/spectral/detail/spectral_util.cuh>
-#include <raft/spectral/eigen_solvers.cuh>
-#include <raft/spectral/matrix_wrappers.hpp>
+#include <tuple>
 
 namespace raft {
 namespace spectral {
diff --git a/cpp/include/raft/spectral/detail/warn_dbg.hpp b/cpp/include/raft/spectral/detail/warn_dbg.hpp
index d714e1dc86..26453e4398 100644
--- a/cpp/include/raft/spectral/detail/warn_dbg.hpp
+++ b/cpp/include/raft/spectral/detail/warn_dbg.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,11 @@
  */
 #pragma once
 
+#include <raft/core/detail/macros.hpp>
+
 #include <stdexcept>
 #include <string>
 
-#include <raft/core/detail/macros.hpp>
-
 #ifdef DEBUG
 #define COUT() (std::cout)
 #define CERR() (std::cerr)
diff --git a/cpp/include/raft/spectral/modularity_maximization.cuh b/cpp/include/raft/spectral/modularity_maximization.cuh
index 7d4f42f952..ab1398a2a1 100644
--- a/cpp/include/raft/spectral/modularity_maximization.cuh
+++ b/cpp/include/raft/spectral/modularity_maximization.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,10 +18,10 @@
 
 #pragma once
 
-#include <tuple>
-
 #include <raft/spectral/detail/modularity_maximization.hpp>
 
+#include <tuple>
+
 namespace raft {
 namespace spectral {
 
diff --git a/cpp/include/raft/spectral/partition.cuh b/cpp/include/raft/spectral/partition.cuh
index b7289aae3d..f7ea456ac5 100644
--- a/cpp/include/raft/spectral/partition.cuh
+++ b/cpp/include/raft/spectral/partition.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,10 +19,10 @@
 
 #pragma once
 
-#include <tuple>
-
 #include <raft/spectral/detail/partition.hpp>
 
+#include <tuple>
+
 namespace raft {
 namespace spectral {
 
diff --git a/cpp/include/raft/stats/detail/adjusted_rand_index.cuh b/cpp/include/raft/stats/detail/adjusted_rand_index.cuh
index 120d4d1686..b108697e51 100644
--- a/cpp/include/raft/stats/detail/adjusted_rand_index.cuh
+++ b/cpp/include/raft/stats/detail/adjusted_rand_index.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,20 +23,23 @@
 #pragma once
 
 #include "contingencyMatrix.cuh"
-#include <cub/cub.cuh>
-#include <math.h>
+
 #include <raft/linalg/map_then_reduce.cuh>
 #include <raft/linalg/reduce.cuh>
 #include <raft/stats/histogram.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
 #include <thrust/device_ptr.h>
 #include <thrust/execution_policy.h>
 #include <thrust/extrema.h>
 
+#include <math.h>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/batched/silhouette_score.cuh b/cpp/include/raft/stats/detail/batched/silhouette_score.cuh
index 559532b0de..643ef77500 100644
--- a/cpp/include/raft/stats/detail/batched/silhouette_score.cuh
+++ b/cpp/include/raft/stats/detail/batched/silhouette_score.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,13 +17,16 @@
 #pragma once
 
 #include "../silhouette_score.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cuda_stream_pool.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/device_atomics.cuh>
+
 #include <rmm/device_uvector.hpp>
 #include <rmm/exec_policy.hpp>
+
 #include <thrust/device_vector.h>
 #include <thrust/fill.h>
 #include <thrust/reduce.h>
diff --git a/cpp/include/raft/stats/detail/contingencyMatrix.cuh b/cpp/include/raft/stats/detail/contingencyMatrix.cuh
index 9f19b798ea..afc4f25a66 100644
--- a/cpp/include/raft/stats/detail/contingencyMatrix.cuh
+++ b/cpp/include/raft/stats/detail/contingencyMatrix.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,13 +19,12 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cub/cub.cuh>
 #include <thrust/device_ptr.h>
 #include <thrust/execution_policy.h>
 #include <thrust/extrema.h>
 #include <thrust/reduce.h>
 
-#include <cub/cub.cuh>
-
 #include <math.h>
 
 namespace raft {
diff --git a/cpp/include/raft/stats/detail/dispersion.cuh b/cpp/include/raft/stats/detail/dispersion.cuh
index d873fcf79a..07b7daf678 100644
--- a/cpp/include/raft/stats/detail/dispersion.cuh
+++ b/cpp/include/raft/stats/detail/dispersion.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,17 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-#include <memory>
 #include <raft/core/interruptible.hpp>
 #include <raft/linalg/eltwise.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
+#include <memory>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/entropy.cuh b/cpp/include/raft/stats/detail/entropy.cuh
index fc4fc5fb6b..4a73cd22a8 100644
--- a/cpp/include/raft/stats/detail/entropy.cuh
+++ b/cpp/include/raft/stats/detail/entropy.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,15 +20,18 @@
  */
 
 #pragma once
-#include <cub/cub.cuh>
-#include <math.h>
 #include <raft/linalg/divide.cuh>
 #include <raft/linalg/map_then_reduce.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
+#include <math.h>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/histogram.cuh b/cpp/include/raft/stats/detail/histogram.cuh
index dee61cf1eb..68c3b03dcc 100644
--- a/cpp/include/raft/stats/detail/histogram.cuh
+++ b/cpp/include/raft/stats/detail/histogram.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,6 +21,7 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/seive.hpp>
 #include <raft/util/vectorized.cuh>
+
 #include <stdint.h>
 
 // This file is a shameless amalgamation of independent works done by
diff --git a/cpp/include/raft/stats/detail/kl_divergence.cuh b/cpp/include/raft/stats/detail/kl_divergence.cuh
index 6d9cc41790..14ff9f7459 100644
--- a/cpp/include/raft/stats/detail/kl_divergence.cuh
+++ b/cpp/include/raft/stats/detail/kl_divergence.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,12 +21,14 @@
 
 #pragma once
 
-#include <math.h>
 #include <raft/linalg/map_then_reduce.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 
+#include <math.h>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/mutual_info_score.cuh b/cpp/include/raft/stats/detail/mutual_info_score.cuh
index cc80d6c752..c67c5b0c6e 100644
--- a/cpp/include/raft/stats/detail/mutual_info_score.cuh
+++ b/cpp/include/raft/stats/detail/mutual_info_score.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -25,16 +25,19 @@
  */
 #pragma once
 
-#include <cub/cub.cuh>
-#include <math.h>
 #include <raft/core/interruptible.hpp>
 #include <raft/linalg/reduce.cuh>
 #include <raft/stats/contingency_matrix.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
+#include <math.h>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/neighborhood_recall.cuh b/cpp/include/raft/stats/detail/neighborhood_recall.cuh
index 3f05225e74..fe3a3f6ece 100644
--- a/cpp/include/raft/stats/detail/neighborhood_recall.cuh
+++ b/cpp/include/raft/stats/detail/neighborhood_recall.cuh
@@ -16,7 +16,6 @@
 
 #pragma once
 
-#include <cstddef>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/error.hpp>
 #include <raft/core/host_mdspan.hpp>
@@ -27,9 +26,9 @@
 #include <raft/core/resources.hpp>
 
 #include <cub/cub.cuh>
-
 #include <cuda/atomic>
 
+#include <cstddef>
 #include <optional>
 
 namespace raft::stats::detail {
diff --git a/cpp/include/raft/stats/detail/rand_index.cuh b/cpp/include/raft/stats/detail/rand_index.cuh
index 744e8dd8e8..4747547e35 100644
--- a/cpp/include/raft/stats/detail/rand_index.cuh
+++ b/cpp/include/raft/stats/detail/rand_index.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -52,13 +52,16 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-#include <math.h>
 #include <raft/core/interruptible.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
+#include <math.h>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/scores.cuh b/cpp/include/raft/stats/detail/scores.cuh
index ae592289ac..947df6848a 100644
--- a/cpp/include/raft/stats/detail/scores.cuh
+++ b/cpp/include/raft/stats/detail/scores.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,6 @@
 
 #pragma once
 
-#include <memory>
 #include <raft/distance/distance.cuh>
 #include <raft/linalg/eltwise.cuh>
 #include <raft/linalg/power.cuh>
@@ -24,13 +23,17 @@
 #include <raft/spatial/knn/knn.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
+
 #include <thrust/count.h>
 #include <thrust/device_ptr.h>
 #include <thrust/execution_policy.h>
 #include <thrust/reduce.h>
 
+#include <memory>
+
 #define N_THREADS 512
 
 namespace raft {
diff --git a/cpp/include/raft/stats/detail/silhouette_score.cuh b/cpp/include/raft/stats/detail/silhouette_score.cuh
index 889a3f5931..4285f84fcc 100644
--- a/cpp/include/raft/stats/detail/silhouette_score.cuh
+++ b/cpp/include/raft/stats/detail/silhouette_score.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,6 @@
 
 #pragma once
 
-#include <algorithm>
-#include <cub/cub.cuh>
-#include <iostream>
-#include <math.h>
-#include <numeric>
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance.cuh>
@@ -32,8 +27,17 @@
 #include <raft/linalg/reduce.cuh>
 #include <raft/linalg/reduce_cols_by_key.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_scalar.hpp>
 
+#include <cub/cub.cuh>
+
+#include <math.h>
+
+#include <algorithm>
+#include <iostream>
+#include <numeric>
+
 namespace raft {
 namespace stats {
 namespace detail {
diff --git a/cpp/include/raft/stats/detail/trustworthiness_score.cuh b/cpp/include/raft/stats/detail/trustworthiness_score.cuh
index ef37b2a034..8b9e5c2cc8 100644
--- a/cpp/include/raft/stats/detail/trustworthiness_score.cuh
+++ b/cpp/include/raft/stats/detail/trustworthiness_score.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,6 +18,7 @@
 #include <raft/distance/distance.cuh>
 #include <raft/matrix/col_wise_sort.cuh>
 #include <raft/spatial/knn/knn.cuh>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/include/raft/stats/dispersion.cuh b/cpp/include/raft/stats/dispersion.cuh
index 216f637c94..ded7c8178b 100644
--- a/cpp/include/raft/stats/dispersion.cuh
+++ b/cpp/include/raft/stats/dispersion.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,11 +19,12 @@
 
 #pragma once
 
-#include <optional>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/detail/dispersion.cuh>
 
+#include <optional>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/include/raft/stats/minmax.cuh b/cpp/include/raft/stats/minmax.cuh
index 5c5ff346a4..d2c410dab1 100644
--- a/cpp/include/raft/stats/minmax.cuh
+++ b/cpp/include/raft/stats/minmax.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,7 +18,6 @@
 
 #pragma once
 
-#include <optional>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/detail/minmax.cuh>
@@ -26,6 +25,7 @@
 #include <raft/util/cudart_utils.hpp>
 
 #include <limits>
+#include <optional>
 
 namespace raft {
 namespace stats {
diff --git a/cpp/include/raft/util/cache.cuh b/cpp/include/raft/util/cache.cuh
index 255dea16bb..902a1f9ae4 100644
--- a/cpp/include/raft/util/cache.cuh
+++ b/cpp/include/raft/util/cache.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,17 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
-
 #include <raft/core/interruptible.hpp>
 #include <raft/core/logger.hpp>
 #include <raft/util/cache_util.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
 #include <cstddef>
 
 namespace raft::cache {
diff --git a/cpp/include/raft/util/cache_util.cuh b/cpp/include/raft/util/cache_util.cuh
index 9f081bb992..df79f8e78b 100644
--- a/cpp/include/raft/util/cache_util.cuh
+++ b/cpp/include/raft/util/cache_util.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,10 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
 #include <raft/util/cuda_utils.cuh>
 
+#include <cub/cub.cuh>
+
 namespace raft {
 namespace cache {
 
diff --git a/cpp/include/raft/util/cuda_data_type.hpp b/cpp/include/raft/util/cuda_data_type.hpp
index acf49730bd..5d7d36405b 100644
--- a/cpp/include/raft/util/cuda_data_type.hpp
+++ b/cpp/include/raft/util/cuda_data_type.hpp
@@ -15,10 +15,12 @@
  */
 #pragma once
 
-#include <cstdint>
 #include <cuda_fp16.hpp>
+
 #include <library_types.h>
 
+#include <cstdint>
+
 namespace raft {
 
 template <typename T>
diff --git a/cpp/include/raft/util/cuda_rt_essentials.hpp b/cpp/include/raft/util/cuda_rt_essentials.hpp
index 77612f97bc..6b221d6700 100644
--- a/cpp/include/raft/util/cuda_rt_essentials.hpp
+++ b/cpp/include/raft/util/cuda_rt_essentials.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -20,9 +20,10 @@
 // The scope is necessarily limited to ensure that compilation times are
 // minimized. Please make sure not to include large / expensive files from here.
 
-#include <cuda_runtime.h>
 #include <raft/core/error.hpp>
 
+#include <cuda_runtime.h>
+
 #include <cstdio>
 
 namespace raft {
diff --git a/cpp/include/raft/util/cuda_utils.cuh b/cpp/include/raft/util/cuda_utils.cuh
index bf46e069e4..08df368bdf 100644
--- a/cpp/include/raft/util/cuda_utils.cuh
+++ b/cpp/include/raft/util/cuda_utils.cuh
@@ -16,15 +16,17 @@
 
 #pragma once
 
+#include <raft/core/cudart_utils.hpp>
+#include <raft/core/math.hpp>
+#include <raft/core/operators.hpp>
+
 #include <cuda_bf16.h>
 #include <cuda_fp16.h>
 #include <math_constants.h>
+
 #include <stdint.h>
-#include <type_traits>
 
-#include <raft/core/cudart_utils.hpp>
-#include <raft/core/math.hpp>
-#include <raft/core/operators.hpp>
+#include <type_traits>
 // For backward compatibility, we include the follow headers. They contain
 // functionality that were previously contained in cuda_utils.cuh
 #include <raft/util/cuda_dev_essentials.cuh>
diff --git a/cpp/include/raft/util/cudart_utils.hpp b/cpp/include/raft/util/cudart_utils.hpp
index 743ffd743c..e5ce15e8a3 100644
--- a/cpp/include/raft/util/cudart_utils.hpp
+++ b/cpp/include/raft/util/cudart_utils.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -25,9 +25,10 @@
 #include <cuda_fp16.h>
 #include <cuda_runtime_api.h>
 
+#include <execinfo.h>
+
 #include <chrono>
 #include <cstdio>
-#include <execinfo.h>
 #include <iomanip>
 #include <iostream>
 #include <memory>
diff --git a/cpp/include/raft/util/cutlass_utils.cuh b/cpp/include/raft/util/cutlass_utils.cuh
index da402c9427..36b9c5de8e 100644
--- a/cpp/include/raft/util/cutlass_utils.cuh
+++ b/cpp/include/raft/util/cutlass_utils.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,10 @@
 
 #pragma once
 
-#include <cutlass/cutlass.h>
 #include <raft/core/error.hpp>
 
+#include <cutlass/cutlass.h>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/util/detail/cub_wrappers.cuh b/cpp/include/raft/util/detail/cub_wrappers.cuh
index 0ce749d9c8..0327d3f9ab 100644
--- a/cpp/include/raft/util/detail/cub_wrappers.cuh
+++ b/cpp/include/raft/util/detail/cub_wrappers.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,10 @@
 
 #pragma once
 
-#include <cub/cub.cuh>
 #include <rmm/device_uvector.hpp>
 
+#include <cub/cub.cuh>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/util/device_atomics.cuh b/cpp/include/raft/util/device_atomics.cuh
index 14856bed8e..6fc6f42ed3 100644
--- a/cpp/include/raft/util/device_atomics.cuh
+++ b/cpp/include/raft/util/device_atomics.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -28,6 +28,7 @@
  */
 
 #include <cooperative_groups.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/include/raft/util/device_loads_stores.cuh b/cpp/include/raft/util/device_loads_stores.cuh
index 780939ee85..2c954ec99a 100644
--- a/cpp/include/raft/util/device_loads_stores.cuh
+++ b/cpp/include/raft/util/device_loads_stores.cuh
@@ -16,11 +16,13 @@
 
 #pragma once
 
-#include <cstdint>  // uintX_t
-#include <cuda_fp16.h>
 #include <raft/core/device_span.hpp>
 #include <raft/util/cuda_dev_essentials.cuh>  // DI
 
+#include <cuda_fp16.h>
+
+#include <cstdint>  // uintX_t
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/util/device_utils.cuh b/cpp/include/raft/util/device_utils.cuh
index 757c60a731..e524325a52 100644
--- a/cpp/include/raft/util/device_utils.cuh
+++ b/cpp/include/raft/util/device_utils.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include <raft/util/cuda_utils.cuh>
+
 #include <utility>  // pair
 
 namespace raft {
diff --git a/cpp/include/raft/util/fast_int_div.cuh b/cpp/include/raft/util/fast_int_div.cuh
index a0cb8f0f53..8ff850b24b 100644
--- a/cpp/include/raft/util/fast_int_div.cuh
+++ b/cpp/include/raft/util/fast_int_div.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include <raft/util/cuda_utils.cuh>
+
 #include <stdint.h>
 
 namespace raft::util {
diff --git a/cpp/include/raft/util/integer_utils.hpp b/cpp/include/raft/util/integer_utils.hpp
index e9a386c426..5224d5ac4c 100644
--- a/cpp/include/raft/util/integer_utils.hpp
+++ b/cpp/include/raft/util/integer_utils.hpp
@@ -1,7 +1,7 @@
 /*
  * Copyright 2019 BlazingDB, Inc.
  *     Copyright 2019 Eyal Rozenberg <eyalroz@blazingdb.com>
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,8 +23,9 @@
  *
  */
 
-#include <limits>
 #include <raft/core/detail/macros.hpp>
+
+#include <limits>
 #include <stdexcept>
 #include <type_traits>
 
diff --git a/cpp/include/raft/util/memory_type_dispatcher.cuh b/cpp/include/raft/util/memory_type_dispatcher.cuh
index 94d838415a..919e4b3484 100644
--- a/cpp/include/raft/util/memory_type_dispatcher.cuh
+++ b/cpp/include/raft/util/memory_type_dispatcher.cuh
@@ -18,6 +18,7 @@
 #include <raft/core/mdbuffer.cuh>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/memory_type.hpp>
+
 #include <type_traits>
 #include <utility>
 
diff --git a/cpp/include/raft/util/reduction.cuh b/cpp/include/raft/util/reduction.cuh
index 362396f9b8..7e897e67f5 100644
--- a/cpp/include/raft/util/reduction.cuh
+++ b/cpp/include/raft/util/reduction.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <stdint.h>
-
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/util/cuda_dev_essentials.cuh>
 #include <raft/util/warp_primitives.cuh>
 
+#include <stdint.h>
+
 namespace raft {
 
 /**
diff --git a/cpp/include/raft/util/seive.hpp b/cpp/include/raft/util/seive.hpp
index ab7c77ac85..509a04c4e6 100644
--- a/cpp/include/raft/util/seive.hpp
+++ b/cpp/include/raft/util/seive.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,6 +16,7 @@
 #pragma once
 
 #include <raft/util/cuda_utils.cuh>
+
 #include <vector>
 
 // Taken from:
diff --git a/cpp/include/raft/util/vectorized.cuh b/cpp/include/raft/util/vectorized.cuh
index 8af2e498e9..8c56554a13 100644
--- a/cpp/include/raft/util/vectorized.cuh
+++ b/cpp/include/raft/util/vectorized.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,10 @@
 
 #pragma once
 
-#include <cuda_fp16.h>
 #include <raft/util/cuda_utils.cuh>
 
+#include <cuda_fp16.h>
+
 namespace raft {
 
 template <typename math_, int VecLen>
diff --git a/cpp/include/raft/util/warp_primitives.cuh b/cpp/include/raft/util/warp_primitives.cuh
index 94fddbe0f3..953c137cdf 100644
--- a/cpp/include/raft/util/warp_primitives.cuh
+++ b/cpp/include/raft/util/warp_primitives.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,12 @@
 
 #pragma once
 
-#include <stdint.h>
-
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/util/cuda_dev_essentials.cuh>
 
+#include <stdint.h>
+
 namespace raft {
 
 /** True CUDA alignment of a type (adapted from CUB) */
diff --git a/cpp/include/raft_runtime/cluster/kmeans.hpp b/cpp/include/raft_runtime/cluster/kmeans.hpp
index d47faf1510..dcc42fe9d3 100644
--- a/cpp/include/raft_runtime/cluster/kmeans.hpp
+++ b/cpp/include/raft_runtime/cluster/kmeans.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,12 @@
  * limitations under the License.
  */
 
+#include <raft/cluster/kmeans_types.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
-#include <raft/cluster/kmeans_types.hpp>
-
 namespace raft::runtime::cluster::kmeans {
 
 /**
diff --git a/cpp/include/raft_runtime/neighbors/cagra.hpp b/cpp/include/raft_runtime/neighbors/cagra.hpp
index 8389929b15..57cb157fbd 100644
--- a/cpp/include/raft_runtime/neighbors/cagra.hpp
+++ b/cpp/include/raft_runtime/neighbors/cagra.hpp
@@ -16,13 +16,13 @@
 
 #pragma once
 
-#include <raft/neighbors/cagra_types.hpp>
-#include <raft/neighbors/ivf_pq_types.hpp>
-#include <string>
-
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_device_accessor.hpp>
 #include <raft/core/mdspan.hpp>
+#include <raft/neighbors/cagra_types.hpp>
+#include <raft/neighbors/ivf_pq_types.hpp>
+
+#include <string>
 
 namespace raft::runtime::neighbors::cagra {
 
diff --git a/cpp/include/raft_runtime/neighbors/ivf_flat.hpp b/cpp/include/raft_runtime/neighbors/ivf_flat.hpp
index 5b8918ec7f..4f65bd6fef 100644
--- a/cpp/include/raft_runtime/neighbors/ivf_flat.hpp
+++ b/cpp/include/raft_runtime/neighbors/ivf_flat.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include <raft/neighbors/ivf_flat_types.hpp>
+
 #include <string>
 
 namespace raft::runtime::neighbors::ivf_flat {
diff --git a/cpp/include/raft_runtime/random/rmat_rectangular_generator.hpp b/cpp/include/raft_runtime/random/rmat_rectangular_generator.hpp
index bf5259aeba..dc9ba9ba7b 100644
--- a/cpp/include/raft_runtime/random/rmat_rectangular_generator.hpp
+++ b/cpp/include/raft_runtime/random/rmat_rectangular_generator.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,11 @@
  * limitations under the License.
  */
 
-#include <cstdint>
-
 #include <raft/core/resources.hpp>
 #include <raft/random/rng_state.hpp>
 
+#include <cstdint>
+
 namespace raft::runtime::random {
 
 /**
diff --git a/cpp/internal/raft_internal/neighbors/naive_knn.cuh b/cpp/internal/raft_internal/neighbors/naive_knn.cuh
index 35d533316f..79206c7a43 100644
--- a/cpp/internal/raft_internal/neighbors/naive_knn.cuh
+++ b/cpp/internal/raft_internal/neighbors/naive_knn.cuh
@@ -23,7 +23,6 @@
 #include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_utils.cuh>
 
-#include <raft/core/resource/cuda_stream.hpp>
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/device_memory_resource.hpp>
diff --git a/cpp/internal/raft_internal/neighbors/refine_helper.cuh b/cpp/internal/raft_internal/neighbors/refine_helper.cuh
index 4a06116877..665ec23d8e 100644
--- a/cpp/internal/raft_internal/neighbors/refine_helper.cuh
+++ b/cpp/internal/raft_internal/neighbors/refine_helper.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,17 +15,17 @@
  */
 #pragma once
 
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft_internal/neighbors/naive_knn.cuh>
-
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/host_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/random/rng.cuh>
 
+#include <raft_internal/neighbors/naive_knn.cuh>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_uvector.hpp>
 
diff --git a/cpp/src/distance/fused_l2_nn.cu b/cpp/src/distance/fused_l2_nn.cu
index 6011aaec29..f29ab08dc1 100644
--- a/cpp/src/distance/fused_l2_nn.cu
+++ b/cpp/src/distance/fused_l2_nn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,11 @@
  * limitations under the License.
  */
 
-#include <cstdint>            // int64_t
 #include <raft/core/kvp.hpp>  // raft::KeyValuePair
 #include <raft/distance/fused_l2_nn-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_distance_fusedL2NNMinReduce(DataT, OutT, IdxT)                   \
   template void raft::distance::fusedL2NNMinReduce<DataT, OutT, IdxT>(OutT * min,         \
                                                                       const DataT* x,     \
diff --git a/cpp/src/matrix/detail/select_k_double_uint32_t.cu b/cpp/src/matrix/detail/select_k_double_uint32_t.cu
index 9aa4e957af..21c954ca46 100644
--- a/cpp/src/matrix/detail/select_k_double_uint32_t.cu
+++ b/cpp/src/matrix/detail/select_k_double_uint32_t.cu
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 
-#include <cstdint>  // uint32_t
 #include <raft/matrix/detail/select_k-inl.cuh>
 
+#include <cstdint>  // uint32_t
+
 #define instantiate_raft_matrix_detail_select_k(T, IdxT)                      \
   template void raft::matrix::detail::select_k(raft::resources const& handle, \
                                                const T* in_val,               \
diff --git a/cpp/src/neighbors/ball_cover.cu b/cpp/src/neighbors/ball_cover.cu
index 0a59060c8e..e9c78f8e7c 100644
--- a/cpp/src/neighbors/ball_cover.cu
+++ b/cpp/src/neighbors/ball_cover.cu
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 
-#include <cstdint>
 #include <raft/neighbors/ball_cover-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_ball_cover(idx_t, value_t, int_t, matrix_idx_t)                 \
   template void raft::neighbors::ball_cover::build_index<idx_t, value_t, int_t, matrix_idx_t>(     \
     raft::resources const& handle,                                                                 \
diff --git a/cpp/src/neighbors/brute_force_fused_l2_knn_float_int64_t.cu b/cpp/src/neighbors/brute_force_fused_l2_knn_float_int64_t.cu
index d482d24772..4269d27225 100644
--- a/cpp/src/neighbors/brute_force_fused_l2_knn_float_int64_t.cu
+++ b/cpp/src/neighbors/brute_force_fused_l2_knn_float_int64_t.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@
  *
  */
 
-#include <cstdint>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_brute_force_fused_l2_knn(            \
   value_t, idx_t, idx_layout, query_layout)                             \
   template void raft::neighbors::brute_force::fused_l2_knn(             \
diff --git a/cpp/src/neighbors/brute_force_knn_index_float.cu b/cpp/src/neighbors/brute_force_knn_index_float.cu
index faf99ceb3c..de94be4c09 100644
--- a/cpp/src/neighbors/brute_force_knn_index_float.cu
+++ b/cpp/src/neighbors/brute_force_knn_index_float.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  * limitations under the License.
  */
 
-#include <cstdint>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 template void raft::neighbors::brute_force::search<float, int>(
   raft::resources const& res,
   const raft::neighbors::brute_force::index<float>& idx,
diff --git a/cpp/src/neighbors/brute_force_knn_int64_t_float_int64_t.cu b/cpp/src/neighbors/brute_force_knn_int64_t_float_int64_t.cu
index a05b387dae..1c08cf8e82 100644
--- a/cpp/src/neighbors/brute_force_knn_int64_t_float_int64_t.cu
+++ b/cpp/src/neighbors/brute_force_knn_int64_t_float_int64_t.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@
  *
  */
 
-#include <cstdint>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_brute_force_knn(                                         \
   idx_t, value_t, matrix_idx, index_layout, search_layout, epilogue_op)                     \
   template void raft::neighbors::brute_force::                                              \
diff --git a/cpp/src/neighbors/brute_force_knn_int64_t_float_uint32_t.cu b/cpp/src/neighbors/brute_force_knn_int64_t_float_uint32_t.cu
index d1b9a4bb43..809cf6eec0 100644
--- a/cpp/src/neighbors/brute_force_knn_int64_t_float_uint32_t.cu
+++ b/cpp/src/neighbors/brute_force_knn_int64_t_float_uint32_t.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@
  *
  */
 
-#include <cstdint>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_brute_force_knn(                                         \
   idx_t, value_t, matrix_idx, index_layout, search_layout, epilogue_op)                     \
   template void raft::neighbors::brute_force::                                              \
diff --git a/cpp/src/neighbors/brute_force_knn_int_float_int.cu b/cpp/src/neighbors/brute_force_knn_int_float_int.cu
index 354f72e3f5..2ffa864dea 100644
--- a/cpp/src/neighbors/brute_force_knn_int_float_int.cu
+++ b/cpp/src/neighbors/brute_force_knn_int_float_int.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@
  *
  */
 
-#include <cstdint>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_brute_force_knn(                                         \
   idx_t, value_t, matrix_idx, index_layout, search_layout, epilogue_op)                     \
   template void raft::neighbors::brute_force::                                              \
diff --git a/cpp/src/neighbors/brute_force_knn_uint32_t_float_uint32_t.cu b/cpp/src/neighbors/brute_force_knn_uint32_t_float_uint32_t.cu
index 2f6bd9b888..dde92765b5 100644
--- a/cpp/src/neighbors/brute_force_knn_uint32_t_float_uint32_t.cu
+++ b/cpp/src/neighbors/brute_force_knn_uint32_t_float_uint32_t.cu
@@ -1,6 +1,6 @@
 
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@
  *
  */
 
-#include <cstdint>
 #include <raft/neighbors/brute_force-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_neighbors_brute_force_knn(                                         \
   idx_t, value_t, matrix_idx, index_layout, search_layout, epilogue_op)                     \
   template void raft::neighbors::brute_force::                                              \
diff --git a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_float_float_int64_t.cu b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_float_float_int64_t.cu
index a1d6cca7d5..def33e493e 100644
--- a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_float_float_int64_t.cu
+++ b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_float_float_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -29,6 +29,8 @@
     const raft::distance::DistanceType metric,                                                  \
     const uint32_t n_probes,                                                                    \
     const uint32_t k,                                                                           \
+    const uint32_t max_samples,                                                                 \
+    const uint32_t* chunk_indices,                                                              \
     const bool select_min,                                                                      \
     IvfSampleFilterT sample_filter,                                                             \
     IdxT* neighbors,                                                                            \
diff --git a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_half_half_int64_t.cu b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_half_half_int64_t.cu
index 3c467a12d8..e96600ee02 100644
--- a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_half_half_int64_t.cu
+++ b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_half_half_int64_t.cu
@@ -31,6 +31,8 @@
     const raft::distance::DistanceType metric,                                                  \
     const uint32_t n_probes,                                                                    \
     const uint32_t k,                                                                           \
+    const uint32_t max_samples,                                                                 \
+    const uint32_t* chunk_indices,                                                              \
     const bool select_min,                                                                      \
     IvfSampleFilterT sample_filter,                                                             \
     IdxT* neighbors,                                                                            \
diff --git a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_int8_t_int32_t_int64_t.cu b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_int8_t_int32_t_int64_t.cu
index 514301562d..13c9d2e283 100644
--- a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_int8_t_int32_t_int64_t.cu
+++ b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_int8_t_int32_t_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -29,6 +29,8 @@
     const raft::distance::DistanceType metric,                                                  \
     const uint32_t n_probes,                                                                    \
     const uint32_t k,                                                                           \
+    const uint32_t max_samples,                                                                 \
+    const uint32_t* chunk_indices,                                                              \
     const bool select_min,                                                                      \
     IvfSampleFilterT sample_filter,                                                             \
     IdxT* neighbors,                                                                            \
diff --git a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_uint8_t_uint32_t_int64_t.cu b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_uint8_t_uint32_t_int64_t.cu
index 32698a8e80..51f02343fc 100644
--- a/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_uint8_t_uint32_t_int64_t.cu
+++ b/cpp/src/neighbors/detail/ivf_flat_interleaved_scan_uint8_t_uint32_t_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -29,6 +29,8 @@
     const raft::distance::DistanceType metric,                                                  \
     const uint32_t n_probes,                                                                    \
     const uint32_t k,                                                                           \
+    const uint32_t max_samples,                                                                 \
+    const uint32_t* chunk_indices,                                                              \
     const bool select_min,                                                                      \
     IvfSampleFilterT sample_filter,                                                             \
     IdxT* neighbors,                                                                            \
diff --git a/cpp/src/raft_runtime/cluster/cluster_cost_double.cu b/cpp/src/raft_runtime/cluster/cluster_cost_double.cu
index ab0299c5d0..abe07c50e7 100644
--- a/cpp/src/raft_runtime/cluster/cluster_cost_double.cu
+++ b/cpp/src/raft_runtime/cluster/cluster_cost_double.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "cluster_cost.cuh"
+
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
diff --git a/cpp/src/raft_runtime/cluster/cluster_cost_float.cu b/cpp/src/raft_runtime/cluster/cluster_cost_float.cu
index 22cfa1b0ff..1634788b62 100644
--- a/cpp/src/raft_runtime/cluster/cluster_cost_float.cu
+++ b/cpp/src/raft_runtime/cluster/cluster_cost_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "cluster_cost.cuh"
+
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
diff --git a/cpp/src/raft_runtime/cluster/update_centroids_double.cu b/cpp/src/raft_runtime/cluster/update_centroids_double.cu
index 18f618eed0..0bc6e33273 100644
--- a/cpp/src/raft_runtime/cluster/update_centroids_double.cu
+++ b/cpp/src/raft_runtime/cluster/update_centroids_double.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "update_centroids.cuh"
+
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
diff --git a/cpp/src/raft_runtime/cluster/update_centroids_float.cu b/cpp/src/raft_runtime/cluster/update_centroids_float.cu
index 31bb42a661..3f98e3fb38 100644
--- a/cpp/src/raft_runtime/cluster/update_centroids_float.cu
+++ b/cpp/src/raft_runtime/cluster/update_centroids_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "update_centroids.cuh"
+
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 
diff --git a/cpp/src/raft_runtime/distance/fused_l2_min_arg.cu b/cpp/src/raft_runtime/distance/fused_l2_min_arg.cu
index 82a225aca4..f421b2fe39 100644
--- a/cpp/src/raft_runtime/distance/fused_l2_min_arg.cu
+++ b/cpp/src/raft_runtime/distance/fused_l2_min_arg.cu
@@ -23,6 +23,7 @@
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/fused_l2_nn.cuh>
 #include <raft/linalg/norm.cuh>
+
 #include <thrust/for_each.h>
 #include <thrust/tuple.h>
 
diff --git a/cpp/src/raft_runtime/neighbors/cagra_build.cu b/cpp/src/raft_runtime/neighbors/cagra_build.cu
index 5fc472c952..9268551b8c 100644
--- a/cpp/src/raft_runtime/neighbors/cagra_build.cu
+++ b/cpp/src/raft_runtime/neighbors/cagra_build.cu
@@ -17,6 +17,7 @@
 #include <raft/neighbors/cagra.cuh>
 #include <raft/neighbors/ivf_pq.cuh>
 #include <raft/neighbors/ivf_pq_types.hpp>
+
 #include <raft_runtime/neighbors/cagra.hpp>
 
 #include <cuda_fp16.h>
diff --git a/cpp/src/raft_runtime/neighbors/cagra_search.cu b/cpp/src/raft_runtime/neighbors/cagra_search.cu
index 93244077f4..b550d2e521 100644
--- a/cpp/src/raft_runtime/neighbors/cagra_search.cu
+++ b/cpp/src/raft_runtime/neighbors/cagra_search.cu
@@ -15,6 +15,7 @@
  */
 
 #include <raft/neighbors/cagra.cuh>
+
 #include <raft_runtime/neighbors/cagra.hpp>
 
 #include <cuda_fp16.h>
diff --git a/cpp/src/raft_runtime/neighbors/cagra_serialize.cu b/cpp/src/raft_runtime/neighbors/cagra_serialize.cu
index f386bcce8e..0677575e23 100644
--- a/cpp/src/raft_runtime/neighbors/cagra_serialize.cu
+++ b/cpp/src/raft_runtime/neighbors/cagra_serialize.cu
@@ -14,16 +14,17 @@
  * limitations under the License.
  */
 
-#include <sstream>
-#include <string>
-
 #include <raft/core/device_resources.hpp>
 #include <raft/neighbors/cagra_serialize.cuh>
 #include <raft/neighbors/cagra_types.hpp>
+
 #include <raft_runtime/neighbors/cagra.hpp>
 
 #include <cuda_fp16.h>
 
+#include <sstream>
+#include <string>
+
 namespace raft::runtime::neighbors::cagra {
 
 #define RAFT_INST_CAGRA_SERIALIZE(DTYPE)                                                      \
diff --git a/cpp/src/raft_runtime/neighbors/hnsw.cpp b/cpp/src/raft_runtime/neighbors/hnsw.cpp
index 1f9e6b0a0b..6eb770abd6 100644
--- a/cpp/src/raft_runtime/neighbors/hnsw.cpp
+++ b/cpp/src/raft_runtime/neighbors/hnsw.cpp
@@ -14,13 +14,14 @@
  * limitations under the License.
  */
 
-#include <filesystem>
 #include <raft/neighbors/hnsw.hpp>
 #include <raft/neighbors/hnsw_serialize.hpp>
 
 #include <raft_runtime/neighbors/cagra.hpp>
 #include <raft_runtime/neighbors/hnsw.hpp>
 
+#include <filesystem>
+
 namespace raft::neighbors::hnsw {
 #define RAFT_INST_HNSW(T)                                                               \
   template <>                                                                           \
diff --git a/cpp/src/raft_runtime/neighbors/ivf_flat_build.cu b/cpp/src/raft_runtime/neighbors/ivf_flat_build.cu
index 7fccb95411..b0c4ec3a49 100644
--- a/cpp/src/raft_runtime/neighbors/ivf_flat_build.cu
+++ b/cpp/src/raft_runtime/neighbors/ivf_flat_build.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <raft/neighbors/ivf_flat.cuh>
+
 #include <raft_runtime/neighbors/ivf_flat.hpp>
 
 namespace raft::runtime::neighbors::ivf_flat {
diff --git a/cpp/src/raft_runtime/neighbors/ivf_flat_search.cu b/cpp/src/raft_runtime/neighbors/ivf_flat_search.cu
index 7a156a4e02..6f26e8fc5f 100644
--- a/cpp/src/raft_runtime/neighbors/ivf_flat_search.cu
+++ b/cpp/src/raft_runtime/neighbors/ivf_flat_search.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <raft/neighbors/ivf_flat.cuh>
+
 #include <raft_runtime/neighbors/ivf_flat.hpp>
 
 namespace raft::runtime::neighbors::ivf_flat {
diff --git a/cpp/src/raft_runtime/neighbors/ivf_flat_serialize.cu b/cpp/src/raft_runtime/neighbors/ivf_flat_serialize.cu
index 049b8b00da..576112f0a4 100644
--- a/cpp/src/raft_runtime/neighbors/ivf_flat_serialize.cu
+++ b/cpp/src/raft_runtime/neighbors/ivf_flat_serialize.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,14 +14,15 @@
  * limitations under the License.
  */
 
-#include <sstream>
-#include <string>
-
 #include <raft/core/device_resources.hpp>
 #include <raft/neighbors/ivf_flat_serialize.cuh>
 #include <raft/neighbors/ivf_flat_types.hpp>
+
 #include <raft_runtime/neighbors/ivf_flat.hpp>
 
+#include <sstream>
+#include <string>
+
 namespace raft::runtime::neighbors::ivf_flat {
 
 #define RAFT_IVF_FLAT_SERIALIZE_INST(DTYPE)                                            \
diff --git a/cpp/src/raft_runtime/neighbors/ivfpq_build.cu b/cpp/src/raft_runtime/neighbors/ivfpq_build.cu
index 65a29429bc..d5c6cd4d28 100644
--- a/cpp/src/raft_runtime/neighbors/ivfpq_build.cu
+++ b/cpp/src/raft_runtime/neighbors/ivfpq_build.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include <raft/neighbors/ivf_pq.cuh>
+
 #include <raft_runtime/neighbors/ivf_pq.hpp>
 
 namespace raft::runtime::neighbors::ivf_pq {
diff --git a/cpp/src/raft_runtime/random/common.cuh b/cpp/src/raft_runtime/random/common.cuh
index f5f8a7c0ae..06309d2037 100644
--- a/cpp/src/raft_runtime/random/common.cuh
+++ b/cpp/src/raft_runtime/random/common.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,6 +16,7 @@
 
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rmat_rectangular_generator.cuh>
+
 #include <raft_runtime/random/rmat_rectangular_generator.hpp>
 
 #define FUNC_DEF(IdxT, ProbT)                                                          \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_eps_pass_euclidean.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_eps_pass_euclidean.cu
index 0d09f88b65..bbc1b55eb2 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_eps_pass_euclidean.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_eps_pass_euclidean.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_eps_pass(                                  \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdist_func)                               \
   template void                                                                            \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_dist.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_dist.cu
index 3681acf245..8c46888e05 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_dist.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_dist.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_euclidean.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_euclidean.cu
index 3fa20779b7..c5a6970a94 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_euclidean.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_euclidean.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_haversine.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_haversine.cu
index 7abc89cc11..19eb4b681b 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_haversine.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_2d_haversine.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_dist.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_dist.cu
index 6251a86867..6f878e1296 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_dist.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_dist.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_euclidean.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_euclidean.cu
index 07b97ac718..e8fbcd0528 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_euclidean.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_euclidean.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_haversine.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_haversine.cu
index 4fc18184b0..57b8c790ef 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_haversine.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_one_3d_haversine.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_one(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_dist.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_dist.cu
index 882496c7d9..6e1cc1e35e 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_dist.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_dist.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_euclidean.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_euclidean.cu
index 0a736d7e13..d1123aff7d 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_euclidean.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_euclidean.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_haversine.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_haversine.cu
index 23aff93966..3e118e2f8f 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_haversine.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_2d_haversine.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_dist.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_dist.cu
index d3ec2b4c65..674f532e6c 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_dist.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_dist.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_euclidean.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_euclidean.cu
index dd9f0e4658..dd3a422baa 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_euclidean.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_euclidean.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_haversine.cu b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_haversine.cu
index 849bbf0f96..0b05ca91a9 100644
--- a/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_haversine.cu
+++ b/cpp/src/spatial/knn/detail/ball_cover/registers_pass_two_3d_haversine.cu
@@ -23,9 +23,10 @@
  *
  */
 
-#include <cstdint>  // int64_t
 #include <raft/spatial/knn/detail/ball_cover/registers-inl.cuh>
 
+#include <cstdint>  // int64_t
+
 #define instantiate_raft_spatial_knn_detail_rbc_low_dim_pass_two(                 \
   Mvalue_idx, Mvalue_t, Mvalue_int, Mmatrix_idx, Mdims, Mdist_func)               \
   template void raft::spatial::knn::detail::                                      \
diff --git a/cpp/src/spatial/knn/detail/fused_l2_knn_int32_t_float.cu b/cpp/src/spatial/knn/detail/fused_l2_knn_int32_t_float.cu
index b73cf31c58..aba67b8a3d 100644
--- a/cpp/src/spatial/knn/detail/fused_l2_knn_int32_t_float.cu
+++ b/cpp/src/spatial/knn/detail/fused_l2_knn_int32_t_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,12 @@
  * limitations under the License.
  */
 
-#include <cstddef>                           // size_t
-#include <cstdint>                           // int_Xt
 #include <raft/distance/distance_types.hpp>  // DistanceType
 #include <raft/spatial/knn/detail/fused_l2_knn-inl.cuh>
 
+#include <cstddef>  // size_t
+#include <cstdint>  // int_Xt
+
 #define instantiate_raft_spatial_knn_detail_fusedL2Knn(Mvalue_idx, Mvalue_t, MusePrevTopKs)  \
   template void raft::spatial::knn::detail::fusedL2Knn<Mvalue_idx, Mvalue_t, MusePrevTopKs>( \
     size_t D,                                                                                \
diff --git a/cpp/src/spatial/knn/detail/fused_l2_knn_int64_t_float.cu b/cpp/src/spatial/knn/detail/fused_l2_knn_int64_t_float.cu
index 35ef37c984..c9d0c49727 100644
--- a/cpp/src/spatial/knn/detail/fused_l2_knn_int64_t_float.cu
+++ b/cpp/src/spatial/knn/detail/fused_l2_knn_int64_t_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,12 @@
  * limitations under the License.
  */
 
-#include <cstddef>                           // size_t
-#include <cstdint>                           // int_Xt
 #include <raft/distance/distance_types.hpp>  // DistanceType
 #include <raft/spatial/knn/detail/fused_l2_knn-inl.cuh>
 
+#include <cstddef>  // size_t
+#include <cstdint>  // int_Xt
+
 #define instantiate_raft_spatial_knn_detail_fusedL2Knn(Mvalue_idx, Mvalue_t, MusePrevTopKs)  \
   template void raft::spatial::knn::detail::fusedL2Knn<Mvalue_idx, Mvalue_t, MusePrevTopKs>( \
     size_t D,                                                                                \
diff --git a/cpp/src/spatial/knn/detail/fused_l2_knn_uint32_t_float.cu b/cpp/src/spatial/knn/detail/fused_l2_knn_uint32_t_float.cu
index ff23d9c41b..e6e9765151 100644
--- a/cpp/src/spatial/knn/detail/fused_l2_knn_uint32_t_float.cu
+++ b/cpp/src/spatial/knn/detail/fused_l2_knn_uint32_t_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,12 @@
  * limitations under the License.
  */
 
-#include <cstddef>                           // size_t
-#include <cstdint>                           // int_Xt
 #include <raft/distance/distance_types.hpp>  // DistanceType
 #include <raft/spatial/knn/detail/fused_l2_knn-inl.cuh>
 
+#include <cstddef>  // size_t
+#include <cstdint>  // int_Xt
+
 #define instantiate_raft_spatial_knn_detail_fusedL2Knn(Mvalue_idx, Mvalue_t, MusePrevTopKs)  \
   template void raft::spatial::knn::detail::fusedL2Knn<Mvalue_idx, Mvalue_t, MusePrevTopKs>( \
     size_t D,                                                                                \
diff --git a/cpp/template/src/ivf_flat_example.cu b/cpp/template/src/ivf_flat_example.cu
index 2b3980f696..60694aea0f 100644
--- a/cpp/template/src/ivf_flat_example.cu
+++ b/cpp/template/src/ivf_flat_example.cu
@@ -15,6 +15,7 @@
  */
 
 #include "common.cuh"
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
diff --git a/cpp/test/CMakeLists.txt b/cpp/test/CMakeLists.txt
index f23ba2ed1a..9ab9604c09 100644
--- a/cpp/test/CMakeLists.txt
+++ b/cpp/test/CMakeLists.txt
@@ -320,6 +320,7 @@ if(BUILD_TESTS)
     test/sparse/sddmm.cu
     test/sparse/sort.cu
     test/sparse/spgemmi.cu
+    test/sparse/spmm.cu
     test/sparse/symmetrize.cu
   )
 
diff --git a/cpp/test/cluster/cluster_solvers.cu b/cpp/test/cluster/cluster_solvers.cu
index e283f6c3c5..cc0a381bbf 100644
--- a/cpp/test/cluster/cluster_solvers.cu
+++ b/cpp/test/cluster/cluster_solvers.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,17 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/device_id.hpp>
 #include <raft/core/resources.hpp>
-
 #include <raft/spectral/cluster_solvers.cuh>
 #include <raft/spectral/modularity_maximization.cuh>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace spectral {
 
diff --git a/cpp/test/cluster/cluster_solvers_deprecated.cu b/cpp/test/cluster/cluster_solvers_deprecated.cu
index a95b013852..954e3e5bb6 100644
--- a/cpp/test/cluster/cluster_solvers_deprecated.cu
+++ b/cpp/test/cluster/cluster_solvers_deprecated.cu
@@ -14,14 +14,15 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
-
 #include <raft/spectral/cluster_solvers_deprecated.cuh>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace spectral {
 
diff --git a/cpp/test/cluster/kmeans.cu b/cpp/test/cluster/kmeans.cu
index ebbe1c46be..33c17c527d 100644
--- a/cpp/test/cluster/kmeans.cu
+++ b/cpp/test/cluster/kmeans.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,21 +15,25 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <optional>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
 
 #include <raft/cluster/kmeans.cuh>
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/make_blobs.cuh>
 #include <raft/stats/adjusted_rand_index.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
 #include <thrust/fill.h>
 
+#include <gtest/gtest.h>
+
+#include <optional>
+#include <vector>
+
 namespace raft {
 
 template <typename T>
diff --git a/cpp/test/cluster/kmeans_balanced.cu b/cpp/test/cluster/kmeans_balanced.cu
index 5b1d764b28..5009eaf122 100644
--- a/cpp/test/cluster/kmeans_balanced.cu
+++ b/cpp/test/cluster/kmeans_balanced.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,22 +15,26 @@
  */
 
 #include "../test_utils.h"
-#include <gtest/gtest.h>
-#include <optional>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
 
 #include <raft/cluster/kmeans_balanced.cuh>
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/handle.hpp>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/random/make_blobs.cuh>
 #include <raft/stats/adjusted_rand_index.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
 #include <thrust/fill.h>
 
+#include <gtest/gtest.h>
+
+#include <optional>
+#include <vector>
+
 /* This test takes advantage of the fact that make_blobs generates balanced clusters.
  * It doesn't currently test whether the algorithm can make balanced clusters with an imbalanced
  * dataset.
diff --git a/cpp/test/cluster/kmeans_find_k.cu b/cpp/test/cluster/kmeans_find_k.cu
index 8f017a0231..8e05ad3695 100644
--- a/cpp/test/cluster/kmeans_find_k.cu
+++ b/cpp/test/cluster/kmeans_find_k.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,17 +15,19 @@
  */
 
 #include "../test_utils.h"
-#include <gtest/gtest.h>
-#include <optional>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <vector>
 
 #include <raft/cluster/kmeans.cuh>
 #include <raft/core/cudart_utils.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/make_blobs.cuh>
 #include <raft/util/cuda_utils.cuh>
 
+#include <gtest/gtest.h>
+
+#include <optional>
+#include <vector>
+
 namespace raft {
 
 template <typename T>
diff --git a/cpp/test/cluster/linkage.cu b/cpp/test/cluster/linkage.cu
index f8ae46cbf9..ba7ed4254e 100644
--- a/cpp/test/cluster/linkage.cu
+++ b/cpp/test/cluster/linkage.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,13 +24,12 @@
 #undef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/transpose.cuh>
 #include <raft/sparse/coo.hpp>
-
-#include <raft/core/device_mdspan.hpp>
 #include <raft/sparse/hierarchy/single_linkage.cuh>
 #include <raft/util/cudart_utils.hpp>
 
diff --git a/cpp/test/cluster/spectral.cu b/cpp/test/cluster/spectral.cu
index 580564ccbc..b8ee611f29 100644
--- a/cpp/test/cluster/spectral.cu
+++ b/cpp/test/cluster/spectral.cu
@@ -16,13 +16,15 @@
 
 #include "../test_utils.cuh"
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/core/handle.hpp>
 #include <raft/spectral/modularity_maximization.cuh>
 #include <raft/spectral/partition.cuh>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace cluster {
 
diff --git a/cpp/test/core/device_resources_manager.cpp b/cpp/test/core/device_resources_manager.cpp
index 335e1cf578..c7c9e175ea 100644
--- a/cpp/test/core/device_resources_manager.cpp
+++ b/cpp/test/core/device_resources_manager.cpp
@@ -25,7 +25,6 @@
 #include <cuda_runtime_api.h>
 
 #include <gtest/gtest.h>
-
 #include <omp.h>
 
 #include <array>
diff --git a/cpp/test/core/device_setter.cpp b/cpp/test/core/device_setter.cpp
index 5a4ff01346..283cf65980 100644
--- a/cpp/test/core/device_setter.cpp
+++ b/cpp/test/core/device_setter.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,12 +13,14 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-#include <cuda_runtime_api.h>
-#include <gtest/gtest.h>
 #include <raft/core/device_setter.hpp>
 #include <raft/core/logger.hpp>
 #include <raft/util/cuda_rt_essentials.hpp>
 
+#include <cuda_runtime_api.h>
+
+#include <gtest/gtest.h>
+
 namespace raft {
 TEST(DeviceSetter, ScopedDevice)
 {
diff --git a/cpp/test/core/interruptible.cu b/cpp/test/core/interruptible.cu
index aa8f16aa3c..e883e74e6f 100644
--- a/cpp/test/core/interruptible.cu
+++ b/cpp/test/core/interruptible.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,18 @@
  * limitations under the License.
  */
 
-#include <cstddef>
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
-#include <omp.h>
 #include <raft/common/nvtx.hpp>
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/interruptible.hpp>
+
 #include <rmm/cuda_stream.hpp>
+
+#include <gtest/gtest.h>
+#include <omp.h>
+
+#include <cstddef>
+#include <iostream>
+#include <memory>
 #include <thread>
 #include <vector>
 
diff --git a/cpp/test/core/logger.cpp b/cpp/test/core/logger.cpp
index d7f001a700..7f31beed71 100644
--- a/cpp/test/core/logger.cpp
+++ b/cpp/test/core/logger.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,8 +19,10 @@
 #undef RAFT_ACTIVE_LEVEL
 #define RAFT_ACTIVE_LEVEL 6
 
-#include <gtest/gtest.h>
 #include <raft/core/logger.hpp>
+
+#include <gtest/gtest.h>
+
 #include <string>
 
 namespace raft {
diff --git a/cpp/test/core/math_device.cu b/cpp/test/core/math_device.cu
index a11a95fff3..fd6cf1b1c6 100644
--- a/cpp/test/core/math_device.cu
+++ b/cpp/test/core/math_device.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,17 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../test_utils.h"
+
 #include <raft/core/math.hpp>
+
 #include <rmm/cuda_stream.hpp>
 #include <rmm/device_scalar.hpp>
 
 #include <cuda/std/type_traits>
 
+#include <gtest/gtest.h>
+
 #ifdef _RAFT_HAS_CUDA
 #include <cuda_bf16.h>
 #include <cuda_fp16.h>
diff --git a/cpp/test/core/math_host.cpp b/cpp/test/core/math_host.cpp
index 5808905713..f7f016adb6 100644
--- a/cpp/test/core/math_host.cpp
+++ b/cpp/test/core/math_host.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,12 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../test_utils.h"
+
 #include <raft/core/math.hpp>
 
+#include <gtest/gtest.h>
+
 TEST(MathHost, Abs)
 {
   // Integer abs
diff --git a/cpp/test/core/mdarray.cu b/cpp/test/core/mdarray.cu
index b0ab36c6e3..6a5cadd0ed 100644
--- a/cpp/test/core/mdarray.cu
+++ b/cpp/test/core/mdarray.cu
@@ -15,28 +15,31 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/thrust_policy.hpp>
 
-#include <gtest/gtest.h>
 #include <raft/core/device_container_policy.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/host_container_policy.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/managed_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/cuda_stream.hpp>
 #include <rmm/device_uvector.hpp>
 #include <rmm/device_vector.hpp>
 #include <rmm/exec_policy.hpp>
+
 #include <thrust/device_vector.h>
 #include <thrust/execution_policy.h>
 #include <thrust/for_each.h>
 #include <thrust/iterator/counting_iterator.h>
 #include <thrust/sequence.h>
 
+#include <gtest/gtest.h>
+
 namespace {
 namespace stdex = std::experimental;
 void check_status(int32_t* d_status, rmm::cuda_stream_view stream)
diff --git a/cpp/test/core/mdbuffer.cu b/cpp/test/core/mdbuffer.cu
index d93d532938..7d5cc6c210 100644
--- a/cpp/test/core/mdbuffer.cu
+++ b/cpp/test/core/mdbuffer.cu
@@ -15,14 +15,17 @@
  */
 
 #include "../test_utils.h"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/managed_mdarray.hpp>
 #include <raft/core/mdbuffer.cuh>
 #include <raft/core/pinned_mdarray.hpp>
+
+#include <gtest/gtest.h>
+
+#include <cstdint>
 #include <utility>
 #include <variant>
 
diff --git a/cpp/test/core/mdspan_copy.cpp b/cpp/test/core/mdspan_copy.cpp
index 2f938e3035..b8aee7f8f5 100644
--- a/cpp/test/core/mdspan_copy.cpp
+++ b/cpp/test/core/mdspan_copy.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,12 @@
  */
 
 #include "../test_utils.h"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/copy.hpp>
+
+#include <gtest/gtest.h>
+
+#include <cstdint>
 #ifndef RAFT_DISABLE_CUDA
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_resources.hpp>
diff --git a/cpp/test/core/mdspan_copy.cu b/cpp/test/core/mdspan_copy.cu
index 95d7d3befd..b68ba38914 100644
--- a/cpp/test/core/mdspan_copy.cu
+++ b/cpp/test/core/mdspan_copy.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.h"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/copy.cuh>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/host_mdarray.hpp>
 
+#include <gtest/gtest.h>
+
+#include <cstdint>
+
 namespace raft {
 TEST(MDSpanCopy, Mdspan3DDeviceDeviceCuda)
 {
diff --git a/cpp/test/core/mdspan_utils.cu b/cpp/test/core/mdspan_utils.cu
index 90ecba1a10..42e5fa605e 100644
--- a/cpp/test/core/mdspan_utils.cu
+++ b/cpp/test/core/mdspan_utils.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,14 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
 #include <raft/core/device_container_policy.hpp>
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/host_container_policy.hpp>
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/resources.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 
 namespace stdex = std::experimental;
diff --git a/cpp/test/core/memory_type.cpp b/cpp/test/core/memory_type.cpp
index cd8aa6bd9e..07ca66ccc1 100644
--- a/cpp/test/core/memory_type.cpp
+++ b/cpp/test/core/memory_type.cpp
@@ -13,9 +13,11 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+#include <raft/core/memory_type.hpp>
+
 #include <cuda_runtime.h>
+
 #include <gtest/gtest.h>
-#include <raft/core/memory_type.hpp>
 
 namespace raft {
 TEST(MemoryType, IsDeviceAccessible)
diff --git a/cpp/test/core/numpy_serializer.cu b/cpp/test/core/numpy_serializer.cu
index 5c562d68f7..e8f12ad706 100644
--- a/cpp/test/core/numpy_serializer.cu
+++ b/cpp/test/core/numpy_serializer.cu
@@ -14,8 +14,6 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include <raft/core/host_mdarray.hpp>
 #include <raft/core/managed_mdspan.hpp>
 #include <raft/core/resources.hpp>
@@ -25,6 +23,8 @@
 #include <thrust/host_vector.h>
 #include <thrust/universal_vector.h>
 
+#include <gtest/gtest.h>
+
 #include <complex>
 #include <cstdint>
 #include <sstream>
diff --git a/cpp/test/core/nvtx.cpp b/cpp/test/core/nvtx.cpp
index e6c29fa3d8..e726f4e1e4 100644
--- a/cpp/test/core/nvtx.cpp
+++ b/cpp/test/core/nvtx.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,8 +14,9 @@
  * limitations under the License.
  */
 #ifdef NVTX_ENABLED
-#include <gtest/gtest.h>
 #include <raft/core/detail/nvtx.hpp>
+
+#include <gtest/gtest.h>
 /**
  * tests for the functionality of generating next color based on string
  * entered in the NVTX Range marker wrappers
diff --git a/cpp/test/core/operators_device.cu b/cpp/test/core/operators_device.cu
index fa4cd25a15..8642ae76a9 100644
--- a/cpp/test/core/operators_device.cu
+++ b/cpp/test/core/operators_device.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,19 @@
  * limitations under the License.
  */
 
-#include <algorithm>
-#include <cmath>
-
-#include <gtest/gtest.h>
-
 #include "../test_utils.cuh"
+
 #include <raft/core/kvp.hpp>
 #include <raft/core/operators.hpp>
+
 #include <rmm/cuda_stream.hpp>
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <cmath>
+
 template <typename OutT, typename OpT, typename... Args>
 RAFT_KERNEL eval_op_on_device_kernel(OutT* out, OpT op, Args... args)
 {
diff --git a/cpp/test/core/operators_host.cpp b/cpp/test/core/operators_host.cpp
index de66fda919..cbb7f060d9 100644
--- a/cpp/test/core/operators_host.cpp
+++ b/cpp/test/core/operators_host.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,16 @@
  * limitations under the License.
  */
 
-#include <algorithm>
-#include <cmath>
-
-#include <gtest/gtest.h>
-
 #include "../test_utils.h"
+
 #include <raft/core/kvp.hpp>
 #include <raft/core/operators.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <cmath>
+
 TEST(OperatorsHost, IdentityOp)
 {
   raft::identity_op op;
diff --git a/cpp/test/core/seive.cu b/cpp/test/core/seive.cu
index 8634abf3be..8b6c202b6a 100644
--- a/cpp/test/core/seive.cu
+++ b/cpp/test/core/seive.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
 #include <raft/util/seive.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace common {
 TEST(Seive, Test)
diff --git a/cpp/test/core/span.cpp b/cpp/test/core/span.cpp
index 1a21b5ff47..78f4ef7277 100644
--- a/cpp/test/core/span.cpp
+++ b/cpp/test/core/span.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,12 @@
  * limitations under the License.
  */
 #include "test_span.hpp"
+
+#include <raft/core/host_span.hpp>
+
 #include <gtest/gtest.h>
+
 #include <numeric>  // iota
-#include <raft/core/host_span.hpp>
 
 namespace raft {
 TEST(Span, DlfConstructors)
diff --git a/cpp/test/core/span.cu b/cpp/test/core/span.cu
index 667c7b19f7..03d8f61f37 100644
--- a/cpp/test/core/span.cu
+++ b/cpp/test/core/span.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,8 +14,7 @@
  * limitations under the License.
  */
 #include "test_span.hpp"
-#include <gtest/gtest.h>
-#include <numeric>  // iota
+
 #include <raft/core/device_span.hpp>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
@@ -28,6 +27,10 @@
 #include <thrust/iterator/counting_iterator.h>
 #include <thrust/memory.h>
 
+#include <gtest/gtest.h>
+
+#include <numeric>  // iota
+
 namespace raft {
 struct TestStatus {
  private:
diff --git a/cpp/test/core/sparse_matrix.cpp b/cpp/test/core/sparse_matrix.cpp
index a5f26a3e53..2b13bbf1ca 100644
--- a/cpp/test/core/sparse_matrix.cpp
+++ b/cpp/test/core/sparse_matrix.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,10 +13,12 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-#include <gtest/gtest.h>
 #include <raft/core/host_coo_matrix.hpp>
 #include <raft/core/host_csr_matrix.hpp>
 #include <raft/core/resources.hpp>
+
+#include <gtest/gtest.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/test/core/sparse_matrix.cu b/cpp/test/core/sparse_matrix.cu
index 3cf0bc98d8..4a1188bd15 100644
--- a/cpp/test/core/sparse_matrix.cu
+++ b/cpp/test/core/sparse_matrix.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,10 +13,12 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-#include <gtest/gtest.h>
 #include <raft/core/device_coo_matrix.hpp>
 #include <raft/core/device_csr_matrix.hpp>
 #include <raft/core/resources.hpp>
+
+#include <gtest/gtest.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/test/core/stream_view.cpp b/cpp/test/core/stream_view.cpp
index 715c53fe21..4148307979 100644
--- a/cpp/test/core/stream_view.cpp
+++ b/cpp/test/core/stream_view.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
 #include <raft/core/cuda_support.hpp>
 #include <raft/core/stream_view.hpp>
+
+#include <gtest/gtest.h>
 #ifndef RAFT_DISABLE_CUDA
 #include <rmm/cuda_stream_view.hpp>
 #endif
diff --git a/cpp/test/core/temporary_device_buffer.cu b/cpp/test/core/temporary_device_buffer.cu
index f61fa826a9..dbcb1ecd44 100644
--- a/cpp/test/core/temporary_device_buffer.cu
+++ b/cpp/test/core/temporary_device_buffer.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,9 +15,9 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/host_mdarray.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/core/temporary_device_buffer.hpp>
 
diff --git a/cpp/test/distance/dist_adj.cu b/cpp/test/distance/dist_adj.cu
index 7c3ec08cf5..a22fb7b1f9 100644
--- a/cpp/test/distance/dist_adj.cu
+++ b/cpp/test/distance/dist_adj.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,18 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+#include "dist_adj.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
-#include "dist_adj.cuh"
+#include <gtest/gtest.h>
 
 namespace raft {
 namespace distance {
diff --git a/cpp/test/distance/dist_adj.cuh b/cpp/test/distance/dist_adj.cuh
index ee4554ff29..2861cb33de 100644
--- a/cpp/test/distance/dist_adj.cuh
+++ b/cpp/test/distance/dist_adj.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "dist_adj_threshold.cuh"
+
 #include <raft/distance/distance.cuh>
 
 #define instantiate_raft_distance_distance(DT, DataT, AccT, OutT, FinalLambda, IdxT)       \
diff --git a/cpp/test/distance/dist_adj_distance_instance.cu b/cpp/test/distance/dist_adj_distance_instance.cu
index d4685d8095..158a5986c2 100644
--- a/cpp/test/distance/dist_adj_distance_instance.cu
+++ b/cpp/test/distance/dist_adj_distance_instance.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,9 +17,11 @@
 #undef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 #include "dist_adj_threshold.cuh"
-#include <cstdint>
+
 #include <raft/distance/distance-inl.cuh>
 
+#include <cstdint>
+
 #define instantiate_raft_distance_distance(DT, DataT, AccT, OutT, FinalLambda, IdxT) \
   template void raft::distance::distance<DT, DataT, AccT, OutT, FinalLambda, IdxT>(  \
     raft::resources const& handle,                                                   \
diff --git a/cpp/test/distance/distance_base.cuh b/cpp/test/distance/distance_base.cuh
index 40e5013d1b..938cd219d0 100644
--- a/cpp/test/distance/distance_base.cuh
+++ b/cpp/test/distance/distance_base.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,18 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <raft/common/nvtx.hpp>  // common::nvtx::range
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/common/nvtx.hpp>         // common::nvtx::range
 #include <raft/core/device_mdspan.hpp>  // make_device_matrix_view
 #include <raft/core/operators.hpp>      // raft::sqrt
-#include <raft/core/resources.hpp>      // raft::resources
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resources.hpp>  // raft::resources
 #include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>  // raft::distance::DistanceType
 #include <raft/random/rng.cuh>
+
 #include <rmm/device_uvector.hpp>  // rmm::device_uvector
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace distance {
 
diff --git a/cpp/test/distance/fused_l2_nn.cu b/cpp/test/distance/fused_l2_nn.cu
index 2a99acd0e4..6fd8f15808 100644
--- a/cpp/test/distance/fused_l2_nn.cu
+++ b/cpp/test/distance/fused_l2_nn.cu
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/kvp.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/fused_l2_nn.cuh>
@@ -24,6 +24,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace distance {
 
diff --git a/cpp/test/distance/gram.cu b/cpp/test/distance/gram.cu
index a9dbd8328f..e911a25ff1 100644
--- a/cpp/test/distance/gram.cu
+++ b/cpp/test/distance/gram.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,20 @@
 
 #include "../test_utils.cuh"
 #include "gram_base.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
+
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/kernels.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft::distance::kernels {
 
 struct GramMatrixInputs {
diff --git a/cpp/test/distance/gram_base.cuh b/cpp/test/distance/gram_base.cuh
index 1862d1a540..170bcb76c1 100644
--- a/cpp/test/distance/gram_base.cuh
+++ b/cpp/test/distance/gram_base.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,17 @@
  * limitations under the License.
  */
 
-#include <iostream>
-#include <memory>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/kernels.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace distance {
 namespace kernels {
diff --git a/cpp/test/distance/masked_nn.cu b/cpp/test/distance/masked_nn.cu
index 67ef1952ca..34fa07c45b 100644
--- a/cpp/test/distance/masked_nn.cu
+++ b/cpp/test/distance/masked_nn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,7 @@
  */
 
 #include "../test_utils.h"
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/kvp.hpp>
@@ -29,6 +28,10 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+
 namespace raft::distance::masked_nn {
 
 // The adjacency pattern determines what distances get computed.
diff --git a/cpp/test/distance/masked_nn_compress_to_bits.cu b/cpp/test/distance/masked_nn_compress_to_bits.cu
index 1563974e56..2512af5e4f 100644
--- a/cpp/test/distance/masked_nn_compress_to_bits.cu
+++ b/cpp/test/distance/masked_nn_compress_to_bits.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,9 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "../test_utils.h"
-#include <cstdio>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/handle.hpp>
@@ -31,6 +29,11 @@
 #include <raft/util/integer_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
+#include <cstdio>
+#include <iostream>
+
 namespace raft::distance::masked_nn::compress_to_bits {
 
 /**
diff --git a/cpp/test/label/label.cu b/cpp/test/label/label.cu
index bda87d423c..4c3479182f 100644
--- a/cpp/test/label/label.cu
+++ b/cpp/test/label/label.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,14 +14,14 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
+#include "../test_utils.cuh"
 
 #include <raft/label/classlabels.cuh>
-
-#include "../test_utils.cuh"
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <vector>
 
diff --git a/cpp/test/label/merge_labels.cu b/cpp/test/label/merge_labels.cu
index 3e12f9171e..e6cb235b0f 100644
--- a/cpp/test/label/merge_labels.cu
+++ b/cpp/test/label/merge_labels.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,20 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/label/merge_labels.cuh>
-
 #include "../test_utils.cuh"
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+#include <raft/label/merge_labels.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
+
 #include <thrust/device_ptr.h>
 
+#include <gtest/gtest.h>
+
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/lap/lap.cu b/cpp/test/lap/lap.cu
index 7576465633..47407f3ea7 100644
--- a/cpp/test/lap/lap.cu
+++ b/cpp/test/lap/lap.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  * Copyright 2020 KETAN DATE & RAKESH NAGI
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
@@ -22,14 +22,15 @@
  *          for the Linear Assignment Problem." Parallel Computing 57 (2016): 52-72.
  *
  */
-#include <gtest/gtest.h>
 #include <raft/core/resource/cuda_stream.hpp>
+#include <raft/solver/linear_assignment.cuh>
 
 #include <rmm/device_uvector.hpp>
 
-#include <iostream>
+#include <gtest/gtest.h>
 #include <omp.h>
-#include <raft/solver/linear_assignment.cuh>
+
+#include <iostream>
 #include <random>
 
 #define PROBLEMSIZE  1000  // Number of rows/columns
diff --git a/cpp/test/linalg/add.cu b/cpp/test/linalg/add.cu
index 668f4c0691..981db432ee 100644
--- a/cpp/test/linalg/add.cu
+++ b/cpp/test/linalg/add.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,14 @@
 
 #include "../test_utils.cuh"
 #include "add.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/add.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/axpy.cu b/cpp/test/linalg/axpy.cu
index 44ec4c1385..f40dae9ff1 100644
--- a/cpp/test/linalg/axpy.cu
+++ b/cpp/test/linalg/axpy.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,16 +13,17 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/axpy.cuh>
-
-#include "../test_utils.cuh"
-#include <gtest/gtest.h>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 // Reference axpy implementation.
diff --git a/cpp/test/linalg/binary_op.cu b/cpp/test/linalg/binary_op.cu
index df8a38bf4e..fd00e72fe8 100644
--- a/cpp/test/linalg/binary_op.cu
+++ b/cpp/test/linalg/binary_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,14 +16,17 @@
 
 #include "../test_utils.cuh"
 #include "binary_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/binary_op.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/binary_op.cuh b/cpp/test/linalg/binary_op.cuh
index 6b502ecf7c..c76b7b606e 100644
--- a/cpp/test/linalg/binary_op.cuh
+++ b/cpp/test/linalg/binary_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include "../test_utils.cuh"
+
 #include <raft/linalg/binary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 
diff --git a/cpp/test/linalg/cholesky_r1.cu b/cpp/test/linalg/cholesky_r1.cu
index 961ab8a7a2..f87e07402f 100644
--- a/cpp/test/linalg/cholesky_r1.cu
+++ b/cpp/test/linalg/cholesky_r1.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,20 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cusolver_dn_handle.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/cholesky_r1_update.cuh>
 #include <raft/linalg/detail/cusolver_wrappers.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
-#include "../test_utils.cuh"
+#include <gtest/gtest.h>
+
 #include <sstream>
 #include <vector>
 namespace raft {
diff --git a/cpp/test/linalg/coalesced_reduction.cu b/cpp/test/linalg/coalesced_reduction.cu
index 6d78788aef..2061f28d36 100644
--- a/cpp/test/linalg/coalesced_reduction.cu
+++ b/cpp/test/linalg/coalesced_reduction.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "reduce.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/coalesced_reduction.cuh>
@@ -24,6 +24,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/divide.cu b/cpp/test/linalg/divide.cu
index 06d98acb0e..14f3ce4ea6 100644
--- a/cpp/test/linalg/divide.cu
+++ b/cpp/test/linalg/divide.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,14 @@
 
 #include "../test_utils.cuh"
 #include "unary_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/divide.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/dot.cu b/cpp/test/linalg/dot.cu
index 34a2e7e7e2..120ee45202 100644
--- a/cpp/test/linalg/dot.cu
+++ b/cpp/test/linalg/dot.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,15 +13,17 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/dot.cuh>
-
-#include "../test_utils.cuh"
-#include <gtest/gtest.h>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 // Reference dot implementation.
diff --git a/cpp/test/linalg/eig.cu b/cpp/test/linalg/eig.cu
index 49990ce300..460b99aaa0 100644
--- a/cpp/test/linalg/eig.cu
+++ b/cpp/test/linalg/eig.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/eig.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/eig_sel.cu b/cpp/test/linalg/eig_sel.cu
index a067cd140e..38e5b5f85e 100644
--- a/cpp/test/linalg/eig_sel.cu
+++ b/cpp/test/linalg/eig_sel.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,12 +17,14 @@
 #if CUDART_VERSION >= 10010
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/eig.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/eltwise.cu b/cpp/test/linalg/eltwise.cu
index 531af71395..b03db48630 100644
--- a/cpp/test/linalg/eltwise.cu
+++ b/cpp/test/linalg/eltwise.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/eltwise.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/gemm_layout.cu b/cpp/test/linalg/gemm_layout.cu
index fa509aaa29..7a764b77d3 100644
--- a/cpp/test/linalg/gemm_layout.cu
+++ b/cpp/test/linalg/gemm_layout.cu
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/gemm.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/gemv.cu b/cpp/test/linalg/gemv.cu
index e90f88e0e1..a842035229 100644
--- a/cpp/test/linalg/gemv.cu
+++ b/cpp/test/linalg/gemv.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/gemv.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/map.cu b/cpp/test/linalg/map.cu
index 9a88460640..1dcc1732a0 100644
--- a/cpp/test/linalg/map.cu
+++ b/cpp/test/linalg/map.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "unary_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/operators.hpp>
@@ -27,6 +27,8 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/map_then_reduce.cu b/cpp/test/linalg/map_then_reduce.cu
index 8e13babec6..9972285450 100644
--- a/cpp/test/linalg/map_then_reduce.cu
+++ b/cpp/test/linalg/map_then_reduce.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <limits>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/map_reduce.cuh>
@@ -24,9 +23,14 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <limits>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/matrix_vector.cu b/cpp/test/linalg/matrix_vector.cu
index f56f51b48b..498ef118bb 100644
--- a/cpp/test/linalg/matrix_vector.cu
+++ b/cpp/test/linalg/matrix_vector.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "matrix_vector_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
@@ -25,6 +25,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/matrix_vector_op.cu b/cpp/test/linalg/matrix_vector_op.cu
index c39e9dd164..cd812327ae 100644
--- a/cpp/test/linalg/matrix_vector_op.cu
+++ b/cpp/test/linalg/matrix_vector_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,15 @@
 
 #include "../test_utils.cuh"
 #include "matrix_vector_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/matrix_vector_op.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
+
+#include <gtest/gtest.h>
+
 #include <type_traits>
 
 namespace raft {
diff --git a/cpp/test/linalg/matrix_vector_op.cuh b/cpp/test/linalg/matrix_vector_op.cuh
index d28e7cc5b5..25d979aa57 100644
--- a/cpp/test/linalg/matrix_vector_op.cuh
+++ b/cpp/test/linalg/matrix_vector_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,6 +15,7 @@
  */
 
 #include "../test_utils.cuh"
+
 #include <raft/linalg/matrix_vector_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 
diff --git a/cpp/test/linalg/mean_squared_error.cu b/cpp/test/linalg/mean_squared_error.cu
index 5d0694f812..7409ee3357 100644
--- a/cpp/test/linalg/mean_squared_error.cu
+++ b/cpp/test/linalg/mean_squared_error.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -13,15 +13,17 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/mean_squared_error.cuh>
-
-#include "../test_utils.cuh"
-#include <gtest/gtest.h>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/multiply.cu b/cpp/test/linalg/multiply.cu
index 0cd890b575..f90832e828 100644
--- a/cpp/test/linalg/multiply.cu
+++ b/cpp/test/linalg/multiply.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,14 @@
 
 #include "../test_utils.cuh"
 #include "unary_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/multiply.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/norm.cu b/cpp/test/linalg/norm.cu
index 888ebe28d7..e4f064949c 100644
--- a/cpp/test/linalg/norm.cu
+++ b/cpp/test/linalg/norm.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/norm.cuh>
@@ -24,6 +24,8 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/normalize.cu b/cpp/test/linalg/normalize.cu
index 63dbd9a19b..252c3906d6 100644
--- a/cpp/test/linalg/normalize.cu
+++ b/cpp/test/linalg/normalize.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/matrix_vector_op.cuh>
@@ -26,6 +26,8 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/power.cu b/cpp/test/linalg/power.cu
index b8e8355fa0..b804c29334 100644
--- a/cpp/test/linalg/power.cu
+++ b/cpp/test/linalg/power.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/power.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/randomized_svd.cu b/cpp/test/linalg/randomized_svd.cu
index 9e1d3df6dc..01ddba6214 100644
--- a/cpp/test/linalg/randomized_svd.cu
+++ b/cpp/test/linalg/randomized_svd.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_resources.hpp>
 #include <raft/linalg/rsvd.cuh>
 #include <raft/linalg/svd.cuh>
@@ -24,6 +24,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/reduce.cu b/cpp/test/linalg/reduce.cu
index 8578fe9637..3ce8e33efd 100644
--- a/cpp/test/linalg/reduce.cu
+++ b/cpp/test/linalg/reduce.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "reduce.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/detail/macros.hpp>
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
@@ -25,6 +25,8 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/reduce.cuh b/cpp/test/linalg/reduce.cuh
index b191fba6c6..73fad14245 100644
--- a/cpp/test/linalg/reduce.cuh
+++ b/cpp/test/linalg/reduce.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,18 @@
 
 #pragma once
 
-#include <cublas_v2.h>
 #include <raft/core/operators.hpp>
 #include <raft/linalg/detail/cublas_wrappers.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/device_ptr.h>
 #include <thrust/memory.h>
 
+#include <cublas_v2.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/reduce_cols_by_key.cu b/cpp/test/linalg/reduce_cols_by_key.cu
index dbd54a350f..80fd06cad4 100644
--- a/cpp/test/linalg/reduce_cols_by_key.cu
+++ b/cpp/test/linalg/reduce_cols_by_key.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/reduce_cols_by_key.cuh>
@@ -23,6 +23,8 @@
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/reduce_rows_by_key.cu b/cpp/test/linalg/reduce_rows_by_key.cu
index 203dbdd5d1..d8cc3f0295 100644
--- a/cpp/test/linalg/reduce_rows_by_key.cu
+++ b/cpp/test/linalg/reduce_rows_by_key.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/reduce_rows_by_key.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/rsvd.cu b/cpp/test/linalg/rsvd.cu
index b5d10d215c..8a2abe5f0d 100644
--- a/cpp/test/linalg/rsvd.cu
+++ b/cpp/test/linalg/rsvd.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,15 +15,18 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/rsvd.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 #include <algorithm>
 
 namespace raft {
diff --git a/cpp/test/linalg/sqrt.cu b/cpp/test/linalg/sqrt.cu
index 06d1b76b78..6717914fdc 100644
--- a/cpp/test/linalg/sqrt.cu
+++ b/cpp/test/linalg/sqrt.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/sqrt.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/strided_reduction.cu b/cpp/test/linalg/strided_reduction.cu
index 06096a74a7..e964769ab3 100644
--- a/cpp/test/linalg/strided_reduction.cu
+++ b/cpp/test/linalg/strided_reduction.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "reduce.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/operators.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/strided_reduction.cuh>
@@ -24,6 +24,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/subtract.cu b/cpp/test/linalg/subtract.cu
index 292abdc07f..29714f82e6 100644
--- a/cpp/test/linalg/subtract.cu
+++ b/cpp/test/linalg/subtract.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/subtract.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/svd.cu b/cpp/test/linalg/svd.cu
index a092b16e22..400eceded2 100644
--- a/cpp/test/linalg/svd.cu
+++ b/cpp/test/linalg/svd.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/init.cuh>
 #include <raft/linalg/svd.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/ternary_op.cu b/cpp/test/linalg/ternary_op.cu
index 7182f52fcd..1ac2725271 100644
--- a/cpp/test/linalg/ternary_op.cu
+++ b/cpp/test/linalg/ternary_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/linalg/ternary_op.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/transpose.cu b/cpp/test/linalg/transpose.cu
index 6579011856..f6857d3ffa 100644
--- a/cpp/test/linalg/transpose.cu
+++ b/cpp/test/linalg/transpose.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,18 +15,18 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/transpose.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <gtest/gtest.h>
-#include <raft/core/device_mdspan.hpp>
-
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/unary_op.cu b/cpp/test/linalg/unary_op.cu
index ba1bdad2f1..d3eb8cc9d3 100644
--- a/cpp/test/linalg/unary_op.cu
+++ b/cpp/test/linalg/unary_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,15 @@
 
 #include "../test_utils.cuh"
 #include "unary_op.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace linalg {
 
diff --git a/cpp/test/linalg/unary_op.cuh b/cpp/test/linalg/unary_op.cuh
index a276024e14..5c1e049f0f 100644
--- a/cpp/test/linalg/unary_op.cuh
+++ b/cpp/test/linalg/unary_op.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,6 +17,7 @@
 #pragma once
 
 #include "../test_utils.cuh"
+
 #include <raft/linalg/unary_op.cuh>
 #include <raft/util/cuda_utils.cuh>
 
diff --git a/cpp/test/matrix/argmax.cu b/cpp/test/matrix/argmax.cu
index 40ca358fe1..cb3fd4a3fb 100644
--- a/cpp/test/matrix/argmax.cu
+++ b/cpp/test/matrix/argmax.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,17 @@
  */
 
 #include "../test_utils.cuh"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/argmax.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
+#include <cstdint>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/argmin.cu b/cpp/test/matrix/argmin.cu
index 29013bbdd8..060b4a78db 100644
--- a/cpp/test/matrix/argmin.cu
+++ b/cpp/test/matrix/argmin.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,17 @@
  */
 
 #include "../test_utils.cuh"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/argmin.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
+#include <cstdint>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/columnSort.cu b/cpp/test/matrix/columnSort.cu
index 1ea9b4ae67..246ed7179d 100644
--- a/cpp/test/matrix/columnSort.cu
+++ b/cpp/test/matrix/columnSort.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <numeric>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/matrix/col_wise_sort.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <numeric>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/diagonal.cu b/cpp/test/matrix/diagonal.cu
index a75e5290ae..c6e1f1a0d2 100644
--- a/cpp/test/matrix/diagonal.cu
+++ b/cpp/test/matrix/diagonal.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
@@ -23,6 +23,8 @@
 #include <raft/matrix/init.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/eye.cu b/cpp/test/matrix/eye.cu
index 33ed8a00ba..9617d68eca 100644
--- a/cpp/test/matrix/eye.cu
+++ b/cpp/test/matrix/eye.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
-
 #include <raft/matrix/diagonal.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft::matrix {
 
 template <typename T>
diff --git a/cpp/test/matrix/gather.cu b/cpp/test/matrix/gather.cu
index b1228f05ca..4c13d0c1e9 100644
--- a/cpp/test/matrix/gather.cu
+++ b/cpp/test/matrix/gather.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/operators.hpp>
@@ -24,8 +24,11 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/itertools.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 
 template <bool Conditional,
diff --git a/cpp/test/matrix/linewise_op.cu b/cpp/test/matrix/linewise_op.cu
index 714119086c..95ed1fae62 100644
--- a/cpp/test/matrix/linewise_op.cu
+++ b/cpp/test/matrix/linewise_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,8 +16,7 @@
 
 #include "../linalg/matrix_vector_op.cuh"
 #include "../test_utils.cuh"
-#include <cuda_profiler_api.h>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/nvtx.hpp>
 #include <raft/core/operators.hpp>
@@ -26,8 +25,13 @@
 #include <raft/matrix/linewise_op.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <cuda_profiler_api.h>
+
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/math.cu b/cpp/test/matrix/math.cu
index 3596e63603..aefe5727c5 100644
--- a/cpp/test/matrix/math.cu
+++ b/cpp/test/matrix/math.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,20 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/power.cuh>
 #include <raft/matrix/ratio.cuh>
 #include <raft/matrix/reciprocal.cuh>
 #include <raft/matrix/sign_flip.cuh>
 #include <raft/matrix/sqrt.cuh>
 #include <raft/matrix/threshold.cuh>
-
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/matrix.cu b/cpp/test/matrix/matrix.cu
index a10280a30c..ccae4266e9 100644
--- a/cpp/test/matrix/matrix.cu
+++ b/cpp/test/matrix/matrix.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,20 +15,22 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
-
 #include <raft/matrix/copy.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/copy.h>
 #include <thrust/device_ptr.h>
 #include <thrust/iterator/counting_iterator.h>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/norm.cu b/cpp/test/matrix/norm.cu
index e370b7ce0c..2b635c5e28 100644
--- a/cpp/test/matrix/norm.cu
+++ b/cpp/test/matrix/norm.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/norm.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/reverse.cu b/cpp/test/matrix/reverse.cu
index 191fc50198..7aa913b059 100644
--- a/cpp/test/matrix/reverse.cu
+++ b/cpp/test/matrix/reverse.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/reverse.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/scatter.cu b/cpp/test/matrix/scatter.cu
index 3a1a40086e..7f478c7b93 100644
--- a/cpp/test/matrix/scatter.cu
+++ b/cpp/test/matrix/scatter.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/cudart_utils.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
@@ -23,12 +23,15 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 #include <thrust/execution_policy.h>
 #include <thrust/random.h>
 #include <thrust/shuffle.h>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 
 template <typename InputIteratorT, typename MapIteratorT, typename OutputIteratorT, typename IdxT>
diff --git a/cpp/test/matrix/select_k.cuh b/cpp/test/matrix/select_k.cuh
index 412a9ae5a2..7f9b7b3fc3 100644
--- a/cpp/test/matrix/select_k.cuh
+++ b/cpp/test/matrix/select_k.cuh
@@ -15,20 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <raft_internal/matrix/select_k.cuh>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/sparse/detail/utils.h>
 #include <raft/util/cudart_utils.hpp>
 
-#include <gtest/gtest.h>
+#include <raft_internal/matrix/select_k.cuh>
 
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/device_memory_resource.hpp>
 
+#include <gtest/gtest.h>
+
 #include <algorithm>
 #include <numeric>
 
diff --git a/cpp/test/matrix/slice.cu b/cpp/test/matrix/slice.cu
index fbf735aaf7..1a6377580e 100644
--- a/cpp/test/matrix/slice.cu
+++ b/cpp/test/matrix/slice.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/slice.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/matrix/triangular.cu b/cpp/test/matrix/triangular.cu
index 00b8a89b02..ce6dfdfef5 100644
--- a/cpp/test/matrix/triangular.cu
+++ b/cpp/test/matrix/triangular.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/init.cuh>
 #include <raft/matrix/triangular.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace matrix {
 
diff --git a/cpp/test/mr/device/buffer.cpp b/cpp/test/mr/device/buffer.cpp
index 447ab77b35..d14aa09b7a 100644
--- a/cpp/test/mr/device/buffer.cpp
+++ b/cpp/test/mr/device/buffer.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/limiting_resource_adaptor.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace mr {
 namespace device {
diff --git a/cpp/test/mr/host/buffer.cpp b/cpp/test/mr/host/buffer.cpp
index 89b185547a..5688ff6376 100644
--- a/cpp/test/mr/host/buffer.cpp
+++ b/cpp/test/mr/host/buffer.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,12 @@
  * limitations under the License.
  */
 
+#include <raft/mr/host/buffer.hpp>
+
 #include <gtest/gtest.h>
+
 #include <iostream>
 #include <memory>
-#include <raft/mr/host/buffer.hpp>
 
 namespace raft {
 namespace mr {
diff --git a/cpp/test/neighbors/ann_brute_force.cuh b/cpp/test/neighbors/ann_brute_force.cuh
index 1cba6bfb39..6370c5ee83 100644
--- a/cpp/test/neighbors/ann_brute_force.cuh
+++ b/cpp/test/neighbors/ann_brute_force.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,40 +18,40 @@
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
 #include "knn_utils.cuh"
+
 #include <raft/core/device_mdarray.hpp>
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdarray.hpp>
+#include <raft/core/logger.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
+#include <raft/distance/distance_types.hpp>
 #include <raft/linalg/map.cuh>
+#include <raft/matrix/detail/select_warpsort.cuh>
+#include <raft/matrix/gather.cuh>
+#include <raft/neighbors/brute_force.cuh>
+#include <raft/neighbors/brute_force_serialize.cuh>
 #include <raft/neighbors/brute_force_types.hpp>
 #include <raft/neighbors/ivf_list.hpp>
 #include <raft/neighbors/sample_filter.cuh>
+#include <raft/random/rng.cuh>
+#include <raft/stats/mean.cuh>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/fast_int_div.cuh>
-#include <thrust/functional.h>
 
 #include <raft_internal/neighbors/naive_knn.cuh>
 
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/logger.hpp>
-#include <raft/distance/distance_types.hpp>
-#include <raft/matrix/detail/select_warpsort.cuh>
-#include <raft/matrix/gather.cuh>
-#include <raft/neighbors/brute_force.cuh>
-#include <raft/neighbors/brute_force_serialize.cuh>
-#include <raft/random/rng.cuh>
-#include <raft/stats/mean.cuh>
-
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_buffer.hpp>
-
-#include <gtest/gtest.h>
-
 #include <rmm/device_uvector.hpp>
+
+#include <thrust/functional.h>
 #include <thrust/sequence.h>
 
+#include <gtest/gtest.h>
+
 #include <cstddef>
 #include <iostream>
 #include <vector>
diff --git a/cpp/test/neighbors/ann_brute_force/test_float.cu b/cpp/test/neighbors/ann_brute_force/test_float.cu
index f618f44b61..f157b5f65c 100644
--- a/cpp/test/neighbors/ann_brute_force/test_float.cu
+++ b/cpp/test/neighbors/ann_brute_force/test_float.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_brute_force.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::brute_force {
 
 using AnnBruteForceTest_float = AnnBruteForceTest<float, float, std::int64_t>;
diff --git a/cpp/test/neighbors/ann_cagra.cuh b/cpp/test/neighbors/ann_cagra.cuh
index 296a5f07fc..a111de0762 100644
--- a/cpp/test/neighbors/ann_cagra.cuh
+++ b/cpp/test/neighbors/ann_cagra.cuh
@@ -19,13 +19,11 @@
 
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <raft_internal/neighbors/naive_knn.cuh>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/device_resources.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/add.cuh>
 #include <raft/neighbors/cagra.cuh>
@@ -34,13 +32,14 @@
 #include <raft/random/rng.cuh>
 #include <raft/util/itertools.hpp>
 
-#include <rmm/device_buffer.hpp>
+#include <raft_internal/neighbors/naive_knn.cuh>
 
-#include <gtest/gtest.h>
+#include <rmm/device_buffer.hpp>
 
+#include <cuda_fp16.h>
 #include <thrust/sequence.h>
 
-#include <cuda_fp16.h>
+#include <gtest/gtest.h>
 
 #include <cstddef>
 #include <iostream>
diff --git a/cpp/test/neighbors/ann_cagra/test_float_int64_t.cu b/cpp/test/neighbors/ann_cagra/test_float_int64_t.cu
index 6f9e8dbd43..ff7e839abf 100644
--- a/cpp/test/neighbors/ann_cagra/test_float_int64_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_float_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,11 +14,11 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 #include "search_kernel_uint64_t.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, float, std::int64_t> AnnCagraTestF_I64;
diff --git a/cpp/test/neighbors/ann_cagra/test_float_uint32_t.cu b/cpp/test/neighbors/ann_cagra/test_float_uint32_t.cu
index 944c2cbc89..7d29ce4f99 100644
--- a/cpp/test/neighbors/ann_cagra/test_float_uint32_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_float_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, float, std::uint32_t> AnnCagraTestF_U32;
diff --git a/cpp/test/neighbors/ann_cagra/test_half_int64_t.cu b/cpp/test/neighbors/ann_cagra/test_half_int64_t.cu
index fdd510bc5d..bcdd95bece 100644
--- a/cpp/test/neighbors/ann_cagra/test_half_int64_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_half_int64_t.cu
@@ -14,11 +14,11 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 #include "search_kernel_uint64_t.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, half, std::int64_t> AnnCagraTestH_I64;
diff --git a/cpp/test/neighbors/ann_cagra/test_half_uint32_t.cu b/cpp/test/neighbors/ann_cagra/test_half_uint32_t.cu
index dea6ae0c23..ec7144f8d0 100644
--- a/cpp/test/neighbors/ann_cagra/test_half_uint32_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_half_uint32_t.cu
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, half, std::uint32_t> AnnCagraTestH_U32;
diff --git a/cpp/test/neighbors/ann_cagra/test_int8_t_uint32_t.cu b/cpp/test/neighbors/ann_cagra/test_int8_t_uint32_t.cu
index 3d9dc76953..b2242d89b1 100644
--- a/cpp/test/neighbors/ann_cagra/test_int8_t_uint32_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_int8_t_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, std::int8_t, std::uint32_t> AnnCagraTestI8_U32;
diff --git a/cpp/test/neighbors/ann_cagra/test_uint8_t_uint32_t.cu b/cpp/test/neighbors/ann_cagra/test_uint8_t_uint32_t.cu
index c5b1b1704b..302b2bec18 100644
--- a/cpp/test/neighbors/ann_cagra/test_uint8_t_uint32_t.cu
+++ b/cpp/test/neighbors/ann_cagra/test_uint8_t_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_cagra.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::cagra {
 
 typedef AnnCagraTest<float, std::uint8_t, std::uint32_t> AnnCagraTestU8_U32;
diff --git a/cpp/test/neighbors/ann_ivf_flat.cuh b/cpp/test/neighbors/ann_ivf_flat.cuh
index 39439d392d..de6af589fa 100644
--- a/cpp/test/neighbors/ann_ivf_flat.cuh
+++ b/cpp/test/neighbors/ann_ivf_flat.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,41 +17,41 @@
 
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
+
 #include <raft/core/device_mdarray.hpp>
+#include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdarray.hpp>
+#include <raft/core/logger.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/mdspan_types.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
-#include <raft/linalg/map.cuh>
-#include <raft/neighbors/ivf_flat_types.hpp>
-#include <raft/neighbors/ivf_list.hpp>
-#include <raft/neighbors/sample_filter.cuh>
-#include <raft/util/cudart_utils.hpp>
-#include <raft/util/fast_int_div.cuh>
-#include <thrust/functional.h>
-
-#include <raft_internal/neighbors/naive_knn.cuh>
-
-#include <raft/core/device_mdspan.hpp>
-#include <raft/core/logger.hpp>
 #include <raft/distance/distance_types.hpp>
+#include <raft/linalg/map.cuh>
 #include <raft/matrix/gather.cuh>
 #include <raft/neighbors/ivf_flat.cuh>
 #include <raft/neighbors/ivf_flat_helpers.cuh>
+#include <raft/neighbors/ivf_flat_types.hpp>
+#include <raft/neighbors/ivf_list.hpp>
+#include <raft/neighbors/sample_filter.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/spatial/knn/ann.cuh>
 #include <raft/spatial/knn/knn.cuh>
 #include <raft/stats/mean.cuh>
+#include <raft/util/cudart_utils.hpp>
+#include <raft/util/fast_int_div.cuh>
+
+#include <raft_internal/neighbors/naive_knn.cuh>
 
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_buffer.hpp>
-
-#include <gtest/gtest.h>
-
 #include <rmm/device_uvector.hpp>
+
+#include <thrust/functional.h>
 #include <thrust/sequence.h>
 
+#include <gtest/gtest.h>
+
 #include <cstddef>
 #include <iostream>
 #include <vector>
@@ -354,7 +354,7 @@ class AnnIVFFlatTest : public ::testing::TestWithParam<AnnIvfFlatInputs<IdxT>> {
       ivf::resize_list(handle_, lists[label], list_device_spec, list_size, 0);
     }
 
-    idx.recompute_internal_state(handle_);
+    helpers::recompute_internal_state(handle_, &idx);
 
     using interleaved_group = Pow2<kIndexGroupSize>;
 
@@ -608,6 +608,23 @@ const std::vector<AnnIvfFlatInputs<int64_t>> inputs = {
   {1000, 100000, 16, 10, 20, 1024, raft::distance::DistanceType::L2Expanded, true},
   {10000, 131072, 8, 10, 20, 1024, raft::distance::DistanceType::L2Expanded, false},
 
+  // various combinations with k>raft::matrix::detail::select::warpsort::kMaxCapacity
+  {1000, 10000, 16, 1024, 40, 1024, raft::distance::DistanceType::L2SqrtExpanded, true},
+  {1000, 10000, 2053, 512, 50, 1024, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {1000, 10000, 2049, 2048, 70, 1024, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {1000, 10000, 16, 4000, 100, 2048, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {10, 10000, 16, 4000, 100, 2048, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {10, 10000, 16, 4000, 120, 2048, raft::distance::DistanceType::L2SqrtExpanded, true},
+  {20, 100000, 16, 257, 20, 1024, raft::distance::DistanceType::L2SqrtExpanded, true},
+  {1000, 100000, 16, 259, 20, 1024, raft::distance::DistanceType::L2Expanded, true, true},
+  {10000, 131072, 8, 280, 20, 1024, raft::distance::DistanceType::InnerProduct, false},
+  {100000, 1024, 32, 257, 64, 64, raft::distance::DistanceType::L2Expanded, false},
+  {100000, 1024, 32, 257, 64, 64, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {100000, 1024, 32, 257, 64, 64, raft::distance::DistanceType::InnerProduct, false},
+  {100000, 1024, 16, 300, 20, 60, raft::distance::DistanceType::L2Expanded, false},
+  {100000, 1024, 16, 500, 20, 60, raft::distance::DistanceType::L2SqrtExpanded, false},
+  {100000, 1024, 16, 700, 20, 60, raft::distance::DistanceType::InnerProduct, false},
+
   // host input data
   {1000, 10000, 16, 10, 40, 1024, raft::distance::DistanceType::L2Expanded, false, true},
   {1000, 10000, 16, 10, 50, 1024, raft::distance::DistanceType::L2Expanded, false, true},
diff --git a/cpp/test/neighbors/ann_ivf_flat/test_float_int64_t.cu b/cpp/test/neighbors/ann_ivf_flat/test_float_int64_t.cu
index 3bfea283e5..2ff17b8536 100644
--- a/cpp/test/neighbors/ann_ivf_flat/test_float_int64_t.cu
+++ b/cpp/test/neighbors/ann_ivf_flat/test_float_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_ivf_flat.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::ivf_flat {
 
 typedef AnnIVFFlatTest<float, float, std::int64_t> AnnIVFFlatTestF;
diff --git a/cpp/test/neighbors/ann_ivf_flat/test_int8_t_int64_t.cu b/cpp/test/neighbors/ann_ivf_flat/test_int8_t_int64_t.cu
index 2f542bd6ec..6fe12506aa 100644
--- a/cpp/test/neighbors/ann_ivf_flat/test_int8_t_int64_t.cu
+++ b/cpp/test/neighbors/ann_ivf_flat/test_int8_t_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_ivf_flat.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::ivf_flat {
 
 typedef AnnIVFFlatTest<float, int8_t, std::int64_t> AnnIVFFlatTestF_int8;
diff --git a/cpp/test/neighbors/ann_ivf_flat/test_uint8_t_int64_t.cu b/cpp/test/neighbors/ann_ivf_flat/test_uint8_t_int64_t.cu
index 7659707089..ab6001c71b 100644
--- a/cpp/test/neighbors/ann_ivf_flat/test_uint8_t_int64_t.cu
+++ b/cpp/test/neighbors/ann_ivf_flat/test_uint8_t_int64_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_ivf_flat.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::ivf_flat {
 
 typedef AnnIVFFlatTest<float, uint8_t, std::int64_t> AnnIVFFlatTestF_uint8;
diff --git a/cpp/test/neighbors/ann_ivf_pq.cuh b/cpp/test/neighbors/ann_ivf_pq.cuh
index e028ef061f..4ebe02027f 100644
--- a/cpp/test/neighbors/ann_ivf_pq.cuh
+++ b/cpp/test/neighbors/ann_ivf_pq.cuh
@@ -17,11 +17,9 @@
 
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <raft_internal/neighbors/naive_knn.cuh>
 
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/map.cuh>
 #include <raft/linalg/map_reduce.cuh>
@@ -32,16 +30,18 @@
 #include <raft/neighbors/sample_filter.cuh>
 #include <raft/random/rng.cuh>
 
+#include <raft_internal/neighbors/naive_knn.cuh>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_buffer.hpp>
 #include <rmm/device_vector.hpp>
 #include <rmm/mr/device/managed_memory_resource.hpp>
 
-#include <gtest/gtest.h>
-
 #include <cub/cub.cuh>
 #include <thrust/sequence.h>
 
+#include <gtest/gtest.h>
+
 #include <algorithm>
 #include <cstddef>
 #include <iostream>
diff --git a/cpp/test/neighbors/ann_ivf_pq/test_float_uint32_t.cu b/cpp/test/neighbors/ann_ivf_pq/test_float_uint32_t.cu
index 5405ddc4a3..a6cfab1f19 100644
--- a/cpp/test/neighbors/ann_ivf_pq/test_float_uint32_t.cu
+++ b/cpp/test/neighbors/ann_ivf_pq/test_float_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -19,6 +19,7 @@
 //
 // TODO: consider removing this test or consider adding an instantiation to the
 // library.
+
 #undef RAFT_EXPLICIT_INSTANTIATE_ONLY
 
 #include "../ann_ivf_pq.cuh"
diff --git a/cpp/test/neighbors/ann_nn_descent.cuh b/cpp/test/neighbors/ann_nn_descent.cuh
index 85574442ad..495af081f1 100644
--- a/cpp/test/neighbors/ann_nn_descent.cuh
+++ b/cpp/test/neighbors/ann_nn_descent.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,12 +18,12 @@
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
 
-#include <raft_internal/neighbors/naive_knn.cuh>
-
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/neighbors/nn_descent.cuh>
 #include <raft/util/itertools.hpp>
 
+#include <raft_internal/neighbors/naive_knn.cuh>
+
 #include <gtest/gtest.h>
 
 #include <cstddef>
diff --git a/cpp/test/neighbors/ann_nn_descent/test_float_uint32_t.cu b/cpp/test/neighbors/ann_nn_descent/test_float_uint32_t.cu
index 6aa503ca4d..ec6d04ad12 100644
--- a/cpp/test/neighbors/ann_nn_descent/test_float_uint32_t.cu
+++ b/cpp/test/neighbors/ann_nn_descent/test_float_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_nn_descent.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::experimental::nn_descent {
 
 typedef AnnNNDescentTest<float, float, std::uint32_t> AnnNNDescentTestF_U32;
diff --git a/cpp/test/neighbors/ann_nn_descent/test_int8_t_uint32_t.cu b/cpp/test/neighbors/ann_nn_descent/test_int8_t_uint32_t.cu
index 863f7edcc0..27fa42d636 100644
--- a/cpp/test/neighbors/ann_nn_descent/test_int8_t_uint32_t.cu
+++ b/cpp/test/neighbors/ann_nn_descent/test_int8_t_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_nn_descent.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::experimental::nn_descent {
 
 typedef AnnNNDescentTest<float, int8_t, std::uint32_t> AnnNNDescentTestI8_U32;
diff --git a/cpp/test/neighbors/ann_nn_descent/test_uint8_t_uint32_t.cu b/cpp/test/neighbors/ann_nn_descent/test_uint8_t_uint32_t.cu
index 1a1b38fc19..3afe79dcc4 100644
--- a/cpp/test/neighbors/ann_nn_descent/test_uint8_t_uint32_t.cu
+++ b/cpp/test/neighbors/ann_nn_descent/test_uint8_t_uint32_t.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,10 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
 #include "../ann_nn_descent.cuh"
 
+#include <gtest/gtest.h>
+
 namespace raft::neighbors::experimental::nn_descent {
 
 typedef AnnNNDescentTest<float, uint8_t, std::uint32_t> AnnNNDescentTestUI8_U32;
diff --git a/cpp/test/neighbors/ann_utils.cuh b/cpp/test/neighbors/ann_utils.cuh
index 18860100f3..afd083d512 100644
--- a/cpp/test/neighbors/ann_utils.cuh
+++ b/cpp/test/neighbors/ann_utils.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,6 +16,8 @@
 
 #pragma once
 
+#include "../test_utils.cuh"
+
 #include <raft/core/device_mdarray.hpp>  // raft::make_device_matrix
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
@@ -24,14 +26,14 @@
 #include <raft/spatial/knn/detail/ann_utils.cuh>
 #include <raft/util/cuda_utils.cuh>
 
+#include <raft_internal/neighbors/naive_knn.cuh>
+
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/device_uvector.hpp>
 #include <rmm/mr/device/device_memory_resource.hpp>
 
-#include <raft_internal/neighbors/naive_knn.cuh>
-
-#include "../test_utils.cuh"
 #include <gtest/gtest.h>
+
 #include <iostream>
 
 namespace raft::neighbors {
diff --git a/cpp/test/neighbors/ball_cover.cu b/cpp/test/neighbors/ball_cover.cu
index 62fbdd6edb..ffd17d6d74 100644
--- a/cpp/test/neighbors/ball_cover.cu
+++ b/cpp/test/neighbors/ball_cover.cu
@@ -16,6 +16,7 @@
 
 #include "../test_utils.cuh"
 #include "spatial_data.h"
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/thrust_policy.hpp>
@@ -32,8 +33,9 @@
 #include <thrust/fill.h>
 #include <thrust/transform.h>
 
-#include <cstdint>
 #include <gtest/gtest.h>
+
+#include <cstdint>
 #include <iostream>
 #include <vector>
 
diff --git a/cpp/test/neighbors/epsilon_neighborhood.cu b/cpp/test/neighbors/epsilon_neighborhood.cu
index 8b35e3ca70..7bf56b18d6 100644
--- a/cpp/test/neighbors/epsilon_neighborhood.cu
+++ b/cpp/test/neighbors/epsilon_neighborhood.cu
@@ -15,8 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <memory>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
@@ -25,8 +24,13 @@
 #include <raft/sparse/convert/csr.cuh>
 #include <raft/spatial/knn/epsilon_neighborhood.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <memory>
+
 namespace raft {
 namespace spatial {
 namespace knn {
diff --git a/cpp/test/neighbors/fused_l2_knn.cu b/cpp/test/neighbors/fused_l2_knn.cu
index 80a01ce568..e4d018aff0 100644
--- a/cpp/test/neighbors/fused_l2_knn.cu
+++ b/cpp/test/neighbors/fused_l2_knn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,16 +16,15 @@
 
 #include "../test_utils.cuh"
 #include "./knn_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/distance/distance.cuh>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/brute_force.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/spatial/knn/knn.cuh>
 
-#include <raft/distance/distance.cuh>
-
 #include <rmm/device_buffer.hpp>
 
 #include <gtest/gtest.h>
diff --git a/cpp/test/neighbors/haversine.cu b/cpp/test/neighbors/haversine.cu
index f1b03a0ab5..55d9fe209e 100644
--- a/cpp/test/neighbors/haversine.cu
+++ b/cpp/test/neighbors/haversine.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/spatial/knn/detail/haversine_distance.cuh>
+
 #include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
+
+#include <iostream>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/neighbors/knn.cu b/cpp/test/neighbors/knn.cu
index 5fe88ee89e..a98de85bda 100644
--- a/cpp/test/neighbors/knn.cu
+++ b/cpp/test/neighbors/knn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,10 +15,10 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/brute_force.cuh>
 
diff --git a/cpp/test/neighbors/knn_utils.cuh b/cpp/test/neighbors/knn_utils.cuh
index b79e7cffe8..3b68ec417e 100644
--- a/cpp/test/neighbors/knn_utils.cuh
+++ b/cpp/test/neighbors/knn_utils.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,14 @@
 
 #pragma once
 
-#include <memory>
-
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
 
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
+#include <memory>
+
 namespace raft::spatial::knn {
 template <typename IdxT, typename DistT, typename compareDist>
 struct idx_dist_pair {
diff --git a/cpp/test/neighbors/refine.cu b/cpp/test/neighbors/refine.cu
index c5865a7618..05e6048e56 100644
--- a/cpp/test/neighbors/refine.cu
+++ b/cpp/test/neighbors/refine.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,11 +16,9 @@
 
 #include "../test_utils.cuh"
 #include "ann_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <raft_internal/neighbors/refine_helper.cuh>
 
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/neighbors/detail/refine.cuh>
@@ -28,6 +26,8 @@
 #include <raft/spatial/knn/ann.cuh>
 #include <raft/util/itertools.hpp>
 
+#include <raft_internal/neighbors/refine_helper.cuh>
+
 #include <rmm/cuda_stream_view.hpp>
 
 #include <gtest/gtest.h>
diff --git a/cpp/test/neighbors/tiled_knn.cu b/cpp/test/neighbors/tiled_knn.cu
index f41a36dde3..d378100711 100644
--- a/cpp/test/neighbors/tiled_knn.cu
+++ b/cpp/test/neighbors/tiled_knn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,10 +17,10 @@
 #include "../test_utils.cuh"
 #include "./ann_utils.cuh"
 #include "./knn_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/logger.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance.cuh>  // raft::distance::pairwise_distance
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/transpose.cuh>
diff --git a/cpp/test/random/make_blobs.cu b/cpp/test/random/make_blobs.cu
index d78fbaf15b..f4b9278bb0 100644
--- a/cpp/test/random/make_blobs.cu
+++ b/cpp/test/random/make_blobs.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,18 @@
  */
 
 #include "../test_utils.cuh"
-#include <cub/cub.cuh>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
-
 #include <raft/random/make_blobs.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cub/cub.cuh>
+
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace random {
 
diff --git a/cpp/test/random/multi_variable_gaussian.cu b/cpp/test/random/multi_variable_gaussian.cu
index 0a1969b5c3..62bad8e543 100644
--- a/cpp/test/random/multi_variable_gaussian.cu
+++ b/cpp/test/random/multi_variable_gaussian.cu
@@ -15,9 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <cmath>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cublas_handle.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/cusolver_dn_handle.hpp>
@@ -26,9 +24,14 @@
 #include <raft/random/multi_variable_gaussian.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <random>
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <cmath>
+#include <iostream>
+#include <random>
+
 // mvg.h takes in column-major matrices (as in Fortran)
 #define IDX2C(i, j, ld) (j * ld + i)
 
diff --git a/cpp/test/random/permute.cu b/cpp/test/random/permute.cu
index f95c8c71c0..b8311c6175 100644
--- a/cpp/test/random/permute.cu
+++ b/cpp/test/random/permute.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/permute.cuh>
 #include <raft/random/rng.cuh>
-
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <algorithm>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/random/rmat_rectangular_generator.cu b/cpp/test/random/rmat_rectangular_generator.cu
index 5a314d0f62..8d668f7a8a 100644
--- a/cpp/test/random/rmat_rectangular_generator.cu
+++ b/cpp/test/random/rmat_rectangular_generator.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,21 +14,22 @@
  * limitations under the License.
  */
 
-#include <cub/cub.cuh>
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <sys/timeb.h>
-#include <vector>
-
 #include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/rmat_rectangular_generator.cuh>
 #include <raft/random/rng.cuh>
-
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cub/cub.cuh>
+
+#include <gtest/gtest.h>
+#include <sys/timeb.h>
+
+#include <vector>
+
 namespace raft {
 namespace random {
 
diff --git a/cpp/test/random/rng.cu b/cpp/test/random/rng.cu
index 8ec35de112..a37f150d4c 100644
--- a/cpp/test/random/rng.cu
+++ b/cpp/test/random/rng.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,21 +14,23 @@
  * limitations under the License.
  */
 
-#include <memory>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <sys/timeb.h>
-
 #include "../test_utils.cuh"
-#include <cub/cub.cuh>
-#include <gtest/gtest.h>
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/stats/stddev.cuh>
-
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cub/cub.cuh>
+
+#include <gtest/gtest.h>
+#include <sys/timeb.h>
+
+#include <memory>
+
 namespace raft {
 namespace random {
 
diff --git a/cpp/test/random/rng_discrete.cu b/cpp/test/random/rng_discrete.cu
index d1293f34ea..0a407a457f 100644
--- a/cpp/test/random/rng_discrete.cu
+++ b/cpp/test/random/rng_discrete.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,17 +15,19 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <cmath>
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/add.cuh>
 #include <raft/linalg/unary_op.cuh>
 #include <raft/random/rng.cuh>
-
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <cmath>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/random/rng_int.cu b/cpp/test/random/rng_int.cu
index 88f1a282b4..28d28f8321 100644
--- a/cpp/test/random/rng_int.cu
+++ b/cpp/test/random/rng_int.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,17 @@
  */
 
 #include "../test_utils.cuh"
-#include <cub/cub.cuh>
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <cub/cub.cuh>
+
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace random {
 
diff --git a/cpp/test/random/sample_without_replacement.cu b/cpp/test/random/sample_without_replacement.cu
index 0d02567ea6..81de89921f 100644
--- a/cpp/test/random/sample_without_replacement.cu
+++ b/cpp/test/random/sample_without_replacement.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/random/sample_without_replacement.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
 #include <set>
 #include <vector>
 
diff --git a/cpp/test/sparse/add.cu b/cpp/test/sparse/add.cu
index 9f0ddd551d..c8f9f546ac 100644
--- a/cpp/test/sparse/add.cu
+++ b/cpp/test/sparse/add.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
+#include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/linalg/add.cuh>
-
-#include "../test_utils.cuh"
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <limits>
 
diff --git a/cpp/test/sparse/convert_coo.cu b/cpp/test/sparse/convert_coo.cu
index f05b6dfb08..a33bec1c0b 100644
--- a/cpp/test/sparse/convert_coo.cu
+++ b/cpp/test/sparse/convert_coo.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,15 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
+#include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/convert/coo.cuh>
 #include <raft/sparse/csr.hpp>
-
 #include <raft/util/cudart_utils.hpp>
 
-#include "../test_utils.cuh"
+#include <gtest/gtest.h>
 
 #include <iostream>
 #include <limits>
diff --git a/cpp/test/sparse/convert_csr.cu b/cpp/test/sparse/convert_csr.cu
index 8be107ef3e..4af792a9ea 100644
--- a/cpp/test/sparse/convert_csr.cu
+++ b/cpp/test/sparse/convert_csr.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cuda_utils.cuh>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/sparse/convert/csr.cuh>
 #include <raft/sparse/coo.hpp>
+#include <raft/util/cuda_utils.cuh>
+
+#include <gtest/gtest.h>
 
 #include <iostream>
 
diff --git a/cpp/test/sparse/csr_row_slice.cu b/cpp/test/sparse/csr_row_slice.cu
index 1438498628..f25287d45a 100644
--- a/cpp/test/sparse/csr_row_slice.cu
+++ b/cpp/test/sparse/csr_row_slice.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,18 @@
  * limitations under the License.
  */
 
-#include <cusparse_v2.h>
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
-#include <raft/util/cudart_utils.hpp>
-
-#include <gtest/gtest.h>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/sparse/op/slice.cuh>
+#include <raft/util/cudart_utils.hpp>
 
 #include <rmm/device_uvector.hpp>
 
-#include "../test_utils.cuh"
+#include <cusparse_v2.h>
+#include <gtest/gtest.h>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/csr_to_dense.cu b/cpp/test/sparse/csr_to_dense.cu
index 47168c91ef..a093e64b73 100644
--- a/cpp/test/sparse/csr_to_dense.cu
+++ b/cpp/test/sparse/csr_to_dense.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,18 @@
  * limitations under the License.
  */
 
-#include <cusparse_v2.h>
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
-#include <raft/util/cudart_utils.hpp>
-
-#include <gtest/gtest.h>
 #include <raft/sparse/convert/dense.cuh>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/util/cudart_utils.hpp>
 
 #include <rmm/device_uvector.hpp>
 
-#include "../test_utils.cuh"
+#include <cusparse_v2.h>
+#include <gtest/gtest.h>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/csr_transpose.cu b/cpp/test/sparse/csr_transpose.cu
index 2379fa2de9..f4d010f62e 100644
--- a/cpp/test/sparse/csr_transpose.cu
+++ b/cpp/test/sparse/csr_transpose.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,16 @@
  * limitations under the License.
  */
 
-#include <cusparse_v2.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <gtest/gtest.h>
+#include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/sparse/linalg/transpose.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include "../test_utils.cuh"
+#include <cusparse_v2.h>
+#include <gtest/gtest.h>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/degree.cu b/cpp/test/sparse/degree.cu
index b5b22ff15c..3449104929 100644
--- a/cpp/test/sparse/degree.cu
+++ b/cpp/test/sparse/degree.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,10 +15,11 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <raft/util/cudart_utils.hpp>
 
 #include <raft/sparse/linalg/degree.cuh>
+#include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
 
 #include <iostream>
 
diff --git a/cpp/test/sparse/dist_coo_spmv.cu b/cpp/test/sparse/dist_coo_spmv.cu
index c729334d00..591cd31e55 100644
--- a/cpp/test/sparse/dist_coo_spmv.cu
+++ b/cpp/test/sparse/dist_coo_spmv.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,21 +14,21 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
+#include "../test_utils.cuh"
 
 #include <raft/core/operators.cuh>
 #include <raft/core/operators.hpp>
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/unary_op.cuh>
+#include <raft/sparse/convert/coo.cuh>
 #include <raft/sparse/detail/cusparse_wrappers.h>
+#include <raft/sparse/distance/detail/coo_spmv.cuh>
 #include <raft/util/cudart_utils.hpp>
-#include <rmm/device_uvector.hpp>
 
-#include <raft/sparse/convert/coo.cuh>
-#include <raft/sparse/distance/detail/coo_spmv.cuh>
+#include <rmm/device_uvector.hpp>
 
-#include "../test_utils.cuh"
+#include <gtest/gtest.h>
 
 #include <type_traits>
 
diff --git a/cpp/test/sparse/distance.cu b/cpp/test/sparse/distance.cu
index 6b4e5c7cfa..a61d212720 100644
--- a/cpp/test/sparse/distance.cu
+++ b/cpp/test/sparse/distance.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <cusparse_v2.h>
+#include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/sparse/detail/cusparse_wrappers.h>
-#include <raft/util/cudart_utils.hpp>
-
 #include <raft/sparse/distance/distance.cuh>
+#include <raft/util/cudart_utils.hpp>
 
-#include "../test_utils.cuh"
+#include <cusparse_v2.h>
+#include <gtest/gtest.h>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/filter.cu b/cpp/test/sparse/filter.cu
index 956bb9c351..5816942d7c 100644
--- a/cpp/test/sparse/filter.cu
+++ b/cpp/test/sparse/filter.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/util/cudart_utils.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/op/filter.cuh>
 #include <raft/sparse/op/sort.cuh>
+#include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
 
 #include <iostream>
 
diff --git a/cpp/test/sparse/gram.cu b/cpp/test/sparse/gram.cu
index ca43aa83b9..3505a3ddf5 100644
--- a/cpp/test/sparse/gram.cu
+++ b/cpp/test/sparse/gram.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,9 +21,7 @@
 
 #include "../distance/gram_base.cuh"
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
+
 #include <raft/distance/distance_types.hpp>
 #include <raft/distance/kernels.cuh>
 #include <raft/random/rng.cuh>
@@ -31,8 +29,14 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 #include <raft/util/itertools.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft::distance::kernels {
 
 /**
diff --git a/cpp/test/sparse/mst.cu b/cpp/test/sparse/mst.cu
index de0694ca10..6efd22967b 100644
--- a/cpp/test/sparse/mst.cu
+++ b/cpp/test/sparse/mst.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,30 +14,25 @@
  * limitations under the License.
  */
 
-#include <bits/stdc++.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
-#include <rmm/device_uvector.hpp>
-#include <vector>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/mst/mst.cuh>
 #include <raft/util/cudart_utils.hpp>
 
 #include <rmm/device_buffer.hpp>
-
-#include <gtest/gtest.h>
-
-#include <bits/stdc++.h>
+#include <rmm/device_uvector.hpp>
 
 #include <thrust/execution_policy.h>
 #include <thrust/memory.h>
 #include <thrust/reduce.h>
 
+#include <bits/stdc++.h>
+#include <gtest/gtest.h>
+
 #include <cstddef>
+#include <iostream>
 #include <vector>
 
 template <typename vertex_t, typename edge_t, typename weight_t>
diff --git a/cpp/test/sparse/neighbors/brute_force.cu b/cpp/test/sparse/neighbors/brute_force.cu
index ed5b92afc2..ed2705a253 100644
--- a/cpp/test/sparse/neighbors/brute_force.cu
+++ b/cpp/test/sparse/neighbors/brute_force.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,16 @@
  * limitations under the License.
  */
 
-#include <cusparse_v2.h>
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "../../test_utils.cuh"
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/sparse/neighbors/knn.cuh>
-
 #include <raft/util/cudart_utils.hpp>
 
+#include <cusparse_v2.h>
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace sparse {
 namespace selection {
diff --git a/cpp/test/sparse/neighbors/cross_component_nn.cu b/cpp/test/sparse/neighbors/cross_component_nn.cu
index 7cadf25e88..13f7ae9de4 100644
--- a/cpp/test/sparse/neighbors/cross_component_nn.cu
+++ b/cpp/test/sparse/neighbors/cross_component_nn.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -24,28 +24,28 @@
 //
 // TODO: edge case testing. Reference: https://github.com/rapidsai/raft/issues/1669
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <cub/cub.cuh>
-
-#include <raft/util/cuda_utils.cuh>
-#include <raft/util/cudart_utils.hpp>
-#include <vector>
-
-#include <raft/sparse/linalg/symmetrize.cuh>
-#include <raft/sparse/mst/mst.cuh>
-#include <raft/sparse/neighbors/knn_graph.cuh>
-#include <raft/sparse/selection/cross_component_nn.cuh>
+#include "../../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
 #include <raft/linalg/transpose.cuh>
 #include <raft/sparse/convert/csr.cuh>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/hierarchy/single_linkage.cuh>
+#include <raft/sparse/linalg/symmetrize.cuh>
+#include <raft/sparse/mst/mst.cuh>
+#include <raft/sparse/neighbors/knn_graph.cuh>
+#include <raft/sparse/selection/cross_component_nn.cuh>
+#include <raft/util/cuda_utils.cuh>
+#include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
-#include "../../test_utils.cuh"
+#include <cub/cub.cuh>
+
+#include <gtest/gtest.h>
+
+#include <vector>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/neighbors/knn_graph.cu b/cpp/test/sparse/neighbors/knn_graph.cu
index 232bd76da1..db610099e9 100644
--- a/cpp/test/sparse/neighbors/knn_graph.cu
+++ b/cpp/test/sparse/neighbors/knn_graph.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,16 @@
  */
 
 #include "../../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
+#include <raft/sparse/coo.hpp>
+#include <raft/sparse/neighbors/knn_graph.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
-#include <raft/sparse/coo.hpp>
-#include <raft/sparse/neighbors/knn_graph.cuh>
+#include <gtest/gtest.h>
 
 #include <iostream>
 
diff --git a/cpp/test/sparse/norm.cu b/cpp/test/sparse/norm.cu
index 10003b1fb6..32ccacd463 100644
--- a/cpp/test/sparse/norm.cu
+++ b/cpp/test/sparse/norm.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/norm_types.hpp>
 #include <raft/sparse/linalg/norm.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <limits>
 
diff --git a/cpp/test/sparse/normalize.cu b/cpp/test/sparse/normalize.cu
index 77ddda225a..1e198d00c4 100644
--- a/cpp/test/sparse/normalize.cu
+++ b/cpp/test/sparse/normalize.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/linalg/norm.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <limits>
 
diff --git a/cpp/test/sparse/reduce.cu b/cpp/test/sparse/reduce.cu
index 7713d96821..f777f4781d 100644
--- a/cpp/test/sparse/reduce.cu
+++ b/cpp/test/sparse/reduce.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,21 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-
 #include "../test_utils.cuh"
-#include <iostream>
-#include <limits>
+
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/op/reduce.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <limits>
+
 namespace raft {
 namespace sparse {
 
diff --git a/cpp/test/sparse/row_op.cu b/cpp/test/sparse/row_op.cu
index 4d15a26b9d..702751f290 100644
--- a/cpp/test/sparse/row_op.cu
+++ b/cpp/test/sparse/row_op.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,15 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
+#include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/sparse/csr.hpp>
 #include <raft/sparse/op/row_op.cuh>
-
-#include "../test_utils.cuh"
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <limits>
 
diff --git a/cpp/test/sparse/sddmm.cu b/cpp/test/sparse/sddmm.cu
index 9323ee8c2b..8ff20581c9 100644
--- a/cpp/test/sparse/sddmm.cu
+++ b/cpp/test/sparse/sddmm.cu
@@ -14,19 +14,20 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-
-#include <iostream>
-#include <limits>
+#include "../test_utils.cuh"
 
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/random/make_blobs.cuh>
 #include <raft/sparse/linalg/sddmm.hpp>
 #include <raft/util/cudart_utils.hpp>
+
 #include <thrust/reduce.h>
 
-#include "../test_utils.cuh"
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <limits>
 
 namespace raft {
 namespace sparse {
diff --git a/cpp/test/sparse/sort.cu b/cpp/test/sparse/sort.cu
index 2e26225e78..2e3464f19f 100644
--- a/cpp/test/sparse/sort.cu
+++ b/cpp/test/sparse/sort.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,13 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
+#include <raft/sparse/op/sort.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <raft/sparse/op/sort.cuh>
+#include <gtest/gtest.h>
 
 #include <iostream>
 #include <memory>
diff --git a/cpp/test/sparse/spectral_matrix.cu b/cpp/test/sparse/spectral_matrix.cu
index 0bed73a722..52f7eff10e 100644
--- a/cpp/test/sparse/spectral_matrix.cu
+++ b/cpp/test/sparse/spectral_matrix.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,15 +14,16 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resource/device_id.hpp>
 #include <raft/core/resources.hpp>
-
 #include <raft/spectral/matrix_wrappers.hpp>
 
+#include <gtest/gtest.h>
+
+#include <iostream>
+#include <memory>
+
 namespace raft {
 namespace spectral {
 namespace matrix {
diff --git a/cpp/test/sparse/spgemmi.cu b/cpp/test/sparse/spgemmi.cu
index 71c18dd9ac..1811f3cef4 100644
--- a/cpp/test/sparse/spgemmi.cu
+++ b/cpp/test/sparse/spgemmi.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,19 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
-#include <raft/core/resource/cuda_stream.hpp>
-#include <raft/core/resource/cusparse_handle.hpp>
-
 #include "../test_utils.cuh"
 
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/cusparse_handle.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/transpose.cuh>
 #include <raft/sparse/detail/cusparse_wrappers.h>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 #include <iostream>
 #include <limits>
 
diff --git a/cpp/test/sparse/spmm.cu b/cpp/test/sparse/spmm.cu
new file mode 100644
index 0000000000..f53e857cd8
--- /dev/null
+++ b/cpp/test/sparse/spmm.cu
@@ -0,0 +1,327 @@
+/*
+ * Copyright (c) 2024, NVIDIA CORPORATION.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include "../test_utils.cuh"
+
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resource/thrust_policy.hpp>
+#include <raft/random/rng.cuh>
+#include <raft/sparse/linalg/spmm.hpp>
+#include <raft/util/cuda_utils.cuh>
+
+#include <thrust/fill.h>
+
+#include <gtest/gtest.h>
+
+namespace raft {
+namespace sparse {
+namespace linalg {
+
+template <typename T>
+struct SpmmInputs {
+  bool trans_x;
+  bool trans_y;
+  int M;
+  int N;
+  int K;
+  int ldy;
+  int ldz;
+  bool row_major;
+  T alpha;
+  T beta;
+  unsigned long long int seed;
+};
+
+template <typename T>
+::std::ostream& operator<<(::std::ostream& os, const SpmmInputs<T>& params)
+{
+  os << " trans_x: " << params.trans_x << " trans_y: " << params.trans_y << " M: " << params.M
+     << ", N: " << params.N << ", K: " << params.K << ", ldy: " << params.ldy
+     << ", ldz: " << params.ldz << ", row_major: " << params.row_major
+     << ", alpha: " << params.alpha << ", beta: " << params.beta;
+  return os;
+}
+
+// Reference GEMM implementation.
+template <typename T>
+RAFT_KERNEL naiveGemm(bool trans_x,
+                      bool trans_y,
+                      int M,
+                      int N,
+                      int K,
+                      T alpha,
+                      T* X,
+                      int ldx,
+                      bool x_row_major,
+                      T* Y,
+                      int ldy,
+                      bool y_row_major,
+                      T beta,
+                      T* Z,
+                      int ldz,
+                      bool z_row_major)
+{
+  int tidx = blockIdx.x * blockDim.x + threadIdx.x;
+  int tidy = blockIdx.y * blockDim.y + threadIdx.y;
+
+  for (int m = tidy; m < M; m += (blockDim.y * gridDim.y)) {
+    for (int n = tidx; n < N; n += (blockDim.x * gridDim.x)) {
+      T temp = T(0.0);
+      for (int k = 0; k < K; k++) {
+        int xIndex = x_row_major != trans_x ? m * ldx + k : m + k * ldx;
+        int yIndex = y_row_major != trans_y ? k * ldy + n : k + n * ldy;
+        temp += X[xIndex] * Y[yIndex];
+      }
+      int zIndex = z_row_major ? m * ldz + n : m + n * ldz;
+      Z[zIndex]  = beta * Z[zIndex] + alpha * temp;
+    }
+  }
+}
+
+template <typename T>
+class SpmmTest : public ::testing::TestWithParam<SpmmInputs<T>> {
+ protected:
+  void SetUp() override
+  {
+    params = ::testing::TestWithParam<SpmmInputs<T>>::GetParam();
+
+    cudaStream_t stream = resource::get_cuda_stream(handle);
+
+    // We compute Z = X * Y and compare against reference result
+    // Dimensions of X : M x K
+    // Dimensions of Y : K x N
+    // Dimensions of Z : M x N
+
+    auto [ldx, ldy, ldz, x_size, y_size, z_size] = getXYZStrides();
+
+    RAFT_CUDA_TRY(cudaMalloc(&X, x_size * sizeof(T)));
+    RAFT_CUDA_TRY(cudaMalloc(&Y, y_size * sizeof(T)));
+    RAFT_CUDA_TRY(cudaMalloc(&Z_ref, z_size * sizeof(T)));
+    RAFT_CUDA_TRY(cudaMalloc(&Z, z_size * sizeof(T)));
+
+    raft::random::RngState r(params.seed);
+    raft::random::uniform(handle, r, X, x_size, T(-10.0), T(10.0));
+    raft::random::uniform(handle, r, Y, y_size, T(-10.0), T(10.0));
+
+    RAFT_CUDA_TRY(
+      cudaMalloc(&X_indptr, ((params.trans_x ? params.K : params.M) + 1) * sizeof(int)));
+    RAFT_CUDA_TRY(cudaMalloc(&X_indices, z_size * sizeof(int)));
+    RAFT_CUDA_TRY(cudaMalloc(&X_data, z_size * sizeof(T)));
+
+    // this will erase all entries below eps and update X accordingly
+    X_nnz = generateCsrFromDense(X,
+                                 params.trans_x ? params.K : params.M,
+                                 params.trans_x ? params.M : params.K,
+                                 X_indptr,
+                                 X_indices,
+                                 X_data);
+
+    resource::sync_stream(handle);
+  }
+
+  std::tuple<int, int, int, size_t, size_t, size_t> getXYZStrides()
+  {
+    // X is always row major
+    int ldx = params.trans_x ? params.M : params.K;
+    int ldy =
+      params.ldy > 0 ? params.ldy : (params.row_major != params.trans_y ? params.N : params.K);
+    int ldz       = params.ldz > 0 ? params.ldz : (params.row_major ? params.N : params.M);
+    size_t x_size = params.M * params.K;
+    size_t y_size = params.row_major != params.trans_y ? params.K * ldy : params.N * ldy;
+    size_t z_size = params.row_major ? params.M * ldz : params.N * ldz;
+
+    return {ldx, ldy, ldz, x_size, y_size, z_size};
+  }
+
+  void runTest()
+  {
+    auto stream = resource::get_cuda_stream(handle);
+
+    auto [ldx, ldy, ldz, x_size, y_size, z_size] = getXYZStrides();
+
+    thrust::fill(resource::get_thrust_policy(handle), Z_ref, Z_ref + z_size, T(0.0));
+    thrust::fill(resource::get_thrust_policy(handle), Z, Z + z_size, T(0.0));
+
+    // create csr matrix view
+    auto X_csr_structure = raft::make_device_compressed_structure_view<int, int, int>(
+      X_indptr,
+      X_indices,
+      params.trans_x ? params.K : params.M,
+      params.trans_x ? params.M : params.K,
+      X_nnz);
+    auto X_csr = raft::device_csr_matrix_view<const T, int, int, int>(
+      raft::device_span<const T>(X_data, X_csr_structure.get_nnz()), X_csr_structure);
+
+    auto y_stride_view =
+      params.row_major
+        ? raft::make_device_strided_matrix_view<const T, int, layout_c_contiguous>(
+            Y, params.trans_y ? params.N : params.K, params.trans_y ? params.K : params.N, ldy)
+        : raft::make_device_strided_matrix_view<const T, int, layout_f_contiguous>(
+            Y, params.trans_y ? params.N : params.K, params.trans_y ? params.K : params.N, ldy);
+
+    auto z_stride_view = params.row_major
+                           ? raft::make_device_strided_matrix_view<T, int, layout_c_contiguous>(
+                               Z, params.M, params.N, ldz)
+                           : raft::make_device_strided_matrix_view<T, int, layout_f_contiguous>(
+                               Z, params.M, params.N, ldz);
+
+    T alpha = params.alpha;
+    T beta  = params.beta;
+
+    dim3 blocks(raft::ceildiv<int>(params.M, 128), raft::ceildiv<int>(params.N, 4), 1);
+    dim3 threads(128, 4, 1);
+    naiveGemm<<<blocks, threads, 0, stream>>>(params.trans_x,
+                                              params.trans_y,
+                                              params.M,
+                                              params.N,
+                                              params.K,
+                                              alpha,
+                                              X,
+                                              ldx,
+                                              true,
+                                              Y,
+                                              ldy,
+                                              params.row_major,
+                                              beta,
+                                              Z_ref,
+                                              ldz,
+                                              params.row_major);
+
+    spmm(
+      handle, params.trans_x, params.trans_y, &alpha, X_csr, y_stride_view, &beta, z_stride_view);
+
+    resource::sync_stream(handle, stream);
+
+    ASSERT_TRUE(raft::devArrMatch(Z_ref, Z, z_size, raft::CompareApprox<T>(1e-3f), stream));
+  }
+
+  void TearDown() override
+  {
+    RAFT_CUDA_TRY(cudaFree(Z_ref));
+    RAFT_CUDA_TRY(cudaFree(Z));
+    RAFT_CUDA_TRY(cudaFree(X));
+    RAFT_CUDA_TRY(cudaFree(Y));
+    RAFT_CUDA_TRY(cudaFree(X_indptr));
+    RAFT_CUDA_TRY(cudaFree(X_indices));
+    RAFT_CUDA_TRY(cudaFree(X_data));
+  }
+
+  int generateCsrFromDense(T* dense, int n_rows, int n_cols, int* indptr, int* indices, T* data)
+  {
+    double eps = 1e-4;
+
+    cudaStream_t stream = resource::get_cuda_stream(handle);
+
+    size_t dense_size = n_rows * n_cols;
+    std::vector<T> dense_host(dense_size);
+    raft::update_host(dense_host.data(), dense, dense_size, stream);
+    resource::sync_stream(handle, stream);
+
+    std::vector<int> indptr_host(n_rows + 1);
+    std::vector<int> indices_host(dense_size);
+    std::vector<T> data_host(dense_size);
+
+    // create csr matrix from dense (with threshold)
+    int nnz = 0;
+    for (int i = 0; i < n_rows; ++i) {
+      indptr_host[i] = nnz;
+      for (int j = 0; j < n_cols; ++j) {
+        T value = dense_host[i * n_cols + j];
+        if (value > eps) {
+          indices_host[nnz] = j;
+          data_host[nnz]    = value;
+          nnz++;
+        } else {
+          dense_host[i * n_cols + j] = T(0.0);
+        }
+      }
+    }
+    indptr_host[n_rows] = nnz;
+
+    raft::update_device(dense, dense_host.data(), dense_size, stream);
+    raft::update_device(indptr, indptr_host.data(), n_rows + 1, stream);
+    raft::update_device(indices, indices_host.data(), nnz, stream);
+    raft::update_device(data, data_host.data(), nnz, stream);
+    resource::sync_stream(handle, stream);
+    return nnz;
+  }
+
+ protected:
+  raft::resources handle;
+  SpmmInputs<T> params;
+
+  T* X     = NULL;
+  T* Y     = NULL;
+  T* Z_ref = NULL;
+  T* Z     = NULL;
+
+  // CSR
+  int* X_indptr  = NULL;
+  int* X_indices = NULL;
+  T* X_data      = NULL;
+  int X_nnz      = 0;
+};
+
+// M / N / K / ldy / ldz / y_row_major / z_row_major / seed
+const std::vector<SpmmInputs<float>> inputsf = {
+  {true, false, 80, 70, 80, 0, 0, true, 1.0, 2.0, 76430ULL},
+  {false, false, 80, 100, 40, 0, 0, true, 3.0, 0.0, 426646ULL},
+  {false, false, 20, 100, 20, 0, 0, false, 2.0, 2.0, 237703ULL},
+  {false, true, 100, 60, 30, 0, 0, false, 1.0, 0.0, 538004ULL},
+  {false, false, 80, 70, 80, 106, 0, true, 1.0, 2.0, 76430ULL},
+  {false, false, 80, 100, 40, 0, 110, true, 3.0, 0.0, 426646ULL},
+  {true, false, 20, 100, 20, 106, 0, false, 2.0, 2.0, 237703ULL},
+  {false, false, 100, 60, 30, 0, 110, false, 1.0, 0.0, 538004ULL},
+  {false, false, 50, 10, 60, 106, 110, true, 1.0, 1.0, 73012ULL},
+  {true, false, 90, 90, 30, 106, 110, true, 1.0, 0.0, 538147ULL},
+  {false, false, 30, 100, 10, 106, 110, false, 1.0, 1.0, 412352ULL},
+  {false, false, 40, 80, 100, 106, 110, false, 1.0, 0.0, 2979410ULL},
+  {true, false, 50, 10, 60, 106, 110, true, 1.0, 1.0, 73012ULL},
+  {true, true, 90, 90, 30, 106, 110, true, 1.0, 0.0, 538147ULL},
+  {false, true, 30, 100, 10, 106, 110, false, 1.0, 1.0, 412352ULL},
+  {true, true, 40, 80, 100, 106, 110, false, 1.0, 0.0, 2979410ULL}};
+
+const std::vector<SpmmInputs<double>> inputsd = {
+  {false, false, 10, 70, 40, 0, 0, true, 1.0, 2.0, 535648ULL},
+  {false, false, 30, 30, 30, 0, 0, true, 3.0, 0.0, 956681ULL},
+  {true, false, 70, 80, 50, 0, 0, false, 2.0, 2.0, 875083ULL},
+  {false, false, 80, 90, 70, 0, 0, false, 1.0, 0.0, 50744ULL},
+  {false, false, 10, 70, 40, 106, 0, true, 1.0, 2.0, 535648ULL},
+  {true, false, 30, 30, 30, 0, 110, true, 3.0, 0.0, 956681ULL},
+  {false, false, 70, 80, 50, 106, 0, false, 2.0, 2.0, 875083ULL},
+  {false, false, 80, 90, 70, 0, 110, false, 1.0, 0.0, 50744ULL},
+  {false, true, 90, 90, 30, 106, 110, true, 1.0, 1.0, 506321ULL},
+  {false, false, 40, 100, 70, 106, 110, true, 1.0, 0.0, 638418ULL},
+  {false, true, 80, 50, 30, 106, 110, false, 1.0, 1.0, 701529ULL},
+  {false, false, 50, 80, 60, 106, 110, false, 1.0, 0.0, 893038ULL},
+  {true, false, 90, 90, 30, 106, 110, true, 1.0, 1.0, 506321ULL},
+  {true, true, 40, 100, 70, 106, 110, true, 1.0, 0.0, 638418ULL},
+  {false, true, 80, 50, 30, 106, 110, false, 1.0, 1.0, 701529ULL},
+  {true, true, 50, 80, 60, 106, 110, false, 1.0, 0.0, 893038ULL}};
+
+typedef SpmmTest<float> SpmmTestF;
+TEST_P(SpmmTestF, Result) { runTest(); }
+
+typedef SpmmTest<double> SpmmTestD;
+TEST_P(SpmmTestD, Result) { runTest(); }
+
+INSTANTIATE_TEST_SUITE_P(SpmmTests, SpmmTestF, ::testing::ValuesIn(inputsf));
+
+INSTANTIATE_TEST_SUITE_P(SpmmTests, SpmmTestD, ::testing::ValuesIn(inputsd));
+
+}  // end namespace linalg
+}  // end namespace sparse
+}  // end namespace raft
diff --git a/cpp/test/sparse/symmetrize.cu b/cpp/test/sparse/symmetrize.cu
index 666396eb5b..e1a74dc40b 100644
--- a/cpp/test/sparse/symmetrize.cu
+++ b/cpp/test/sparse/symmetrize.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,18 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
+#include "../test_utils.cuh"
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/sparse/convert/coo.cuh>
 #include <raft/sparse/coo.hpp>
 #include <raft/sparse/linalg/symmetrize.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 #include <rmm/device_uvector.hpp>
 
-#include "../test_utils.cuh"
+#include <gtest/gtest.h>
 
 #include <iostream>
 
diff --git a/cpp/test/stats/accuracy.cu b/cpp/test/stats/accuracy.cu
index 5bc0506e7f..5fa2b3b6de 100644
--- a/cpp/test/stats/accuracy.cu
+++ b/cpp/test/stats/accuracy.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <optional>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/accuracy.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
+
+#include <optional>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/stats/adjusted_rand_index.cu b/cpp/test/stats/adjusted_rand_index.cu
index fb7b3825fc..f2207dfd21 100644
--- a/cpp/test/stats/adjusted_rand_index.cu
+++ b/cpp/test/stats/adjusted_rand_index.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/adjusted_rand_index.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/completeness_score.cu b/cpp/test/stats/completeness_score.cu
index c5c134418c..b985799b0f 100644
--- a/cpp/test/stats/completeness_score.cu
+++ b/cpp/test/stats/completeness_score.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,14 +14,17 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/completeness_score.cuh>
 #include <raft/stats/entropy.cuh>
 #include <raft/stats/mutual_info_score.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/contingencyMatrix.cu b/cpp/test/stats/contingencyMatrix.cu
index acfd1aecfe..221c4b4613 100644
--- a/cpp/test/stats/contingencyMatrix.cu
+++ b/cpp/test/stats/contingencyMatrix.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/contingency_matrix.cuh>
 #include <raft/util/cudart_utils.hpp>
-#include <random>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
+#include <random>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/cov.cu b/cpp/test/stats/cov.cu
index ca9d437717..41812979b6 100644
--- a/cpp/test/stats/cov.cu
+++ b/cpp/test/stats/cov.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,14 +15,17 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/cov.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/dispersion.cu b/cpp/test/stats/dispersion.cu
index 9ef678050d..54b1cee4dd 100644
--- a/cpp/test/stats/dispersion.cu
+++ b/cpp/test/stats/dispersion.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <optional>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/dispersion.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
+
+#include <optional>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/stats/entropy.cu b/cpp/test/stats/entropy.cu
index dea8828b26..54a1e88dcd 100644
--- a/cpp/test/stats/entropy.cu
+++ b/cpp/test/stats/entropy.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,16 +14,20 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/entropy.cuh>
 #include <raft/util/cudart_utils.hpp>
-#include <random>
+
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
+#include <random>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/histogram.cu b/cpp/test/stats/histogram.cu
index 86f708db62..18b8436a73 100644
--- a/cpp/test/stats/histogram.cu
+++ b/cpp/test/stats/histogram.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
@@ -24,6 +24,8 @@
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/homogeneity_score.cu b/cpp/test/stats/homogeneity_score.cu
index 88247f5b50..bcce94fca0 100644
--- a/cpp/test/stats/homogeneity_score.cu
+++ b/cpp/test/stats/homogeneity_score.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,16 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/homogeneity_score.cuh>
 #include <raft/stats/mutual_info_score.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/information_criterion.cu b/cpp/test/stats/information_criterion.cu
index 9e57f2c84f..ac3dff1d7a 100644
--- a/cpp/test/stats/information_criterion.cu
+++ b/cpp/test/stats/information_criterion.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,12 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
-
-#include <raft/stats/information_criterion.cuh>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
+#include <raft/stats/information_criterion.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/device_uvector.hpp>
 
 #include <gtest/gtest.h>
diff --git a/cpp/test/stats/kl_divergence.cu b/cpp/test/stats/kl_divergence.cu
index 5714583675..477fd6d892 100644
--- a/cpp/test/stats/kl_divergence.cu
+++ b/cpp/test/stats/kl_divergence.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,12 +14,15 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/kl_divergence.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/mean.cu b/cpp/test/stats/mean.cu
index 0cb90b6d46..67931e2eed 100644
--- a/cpp/test/stats/mean.cu
+++ b/cpp/test/stats/mean.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,12 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
 
diff --git a/cpp/test/stats/mean_center.cu b/cpp/test/stats/mean_center.cu
index 4d797e60fe..b44d87d1bd 100644
--- a/cpp/test/stats/mean_center.cu
+++ b/cpp/test/stats/mean_center.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,13 +16,15 @@
 
 #include "../linalg/matrix_vector_op.cuh"
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/mean.cuh>
 #include <raft/stats/mean_center.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/meanvar.cu b/cpp/test/stats/meanvar.cu
index df3d9d9c00..725e773cb0 100644
--- a/cpp/test/stats/meanvar.cu
+++ b/cpp/test/stats/meanvar.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/math.cuh>
 #include <raft/random/rng.cuh>
 #include <raft/stats/meanvar.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 #include <algorithm>
 
 namespace raft {
diff --git a/cpp/test/stats/minmax.cu b/cpp/test/stats/minmax.cu
index 3715bc5bd3..7563cb12be 100644
--- a/cpp/test/stats/minmax.cu
+++ b/cpp/test/stats/minmax.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,8 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <limits>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
@@ -24,9 +23,13 @@
 #include <raft/stats/minmax.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
 
+#include <limits>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/mutual_info_score.cu b/cpp/test/stats/mutual_info_score.cu
index 9f31350844..811bc793cf 100644
--- a/cpp/test/stats/mutual_info_score.cu
+++ b/cpp/test/stats/mutual_info_score.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,16 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/stats/mutual_info_score.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/neighborhood_recall.cu b/cpp/test/stats/neighborhood_recall.cu
index 43ae7059bd..1a76154e2e 100644
--- a/cpp/test/stats/neighborhood_recall.cu
+++ b/cpp/test/stats/neighborhood_recall.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -21,12 +21,11 @@
 #include <raft/core/mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
-
-#include <raft_internal/neighbors/naive_knn.cuh>
-
 #include <raft/stats/neighborhood_recall.cuh>
 #include <raft/util/itertools.hpp>
 
+#include <raft_internal/neighbors/naive_knn.cuh>
+
 #include <gtest/gtest.h>
 
 namespace raft::stats {
diff --git a/cpp/test/stats/r2_score.cu b/cpp/test/stats/r2_score.cu
index aa4f069f09..a4a3ed4de6 100644
--- a/cpp/test/stats/r2_score.cu
+++ b/cpp/test/stats/r2_score.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <optional>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/r2_score.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
+
+#include <optional>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/stats/rand_index.cu b/cpp/test/stats/rand_index.cu
index 41b0823e76..678bc455f0 100644
--- a/cpp/test/stats/rand_index.cu
+++ b/cpp/test/stats/rand_index.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,16 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
+#include <raft/core/resources.hpp>
+#include <raft/stats/rand_index.cuh>
 #include <raft/util/cudart_utils.hpp>
 
 #include <gtest/gtest.h>
 
 #include <algorithm>
 #include <iostream>
-#include <raft/core/resources.hpp>
-#include <raft/stats/rand_index.cuh>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/regression_metrics.cu b/cpp/test/stats/regression_metrics.cu
index b0c4cca530..d03784f834 100644
--- a/cpp/test/stats/regression_metrics.cu
+++ b/cpp/test/stats/regression_metrics.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,17 +15,21 @@
  */
 
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <optional>
+
 #include <raft/core/interruptible.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/regression_metrics.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
 #include <stdio.h>
 #include <stdlib.h>
+
+#include <algorithm>
+#include <optional>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/stats/silhouette_score.cu b/cpp/test/stats/silhouette_score.cu
index ed69b2f3ac..ad080f5894 100644
--- a/cpp/test/stats/silhouette_score.cu
+++ b/cpp/test/stats/silhouette_score.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,20 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance_types.hpp>
+#include <raft/stats/silhouette_score.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <raft/stats/silhouette_score.cuh>
-#include <random>
 #include <rmm/device_uvector.hpp>
 
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
+#include <random>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/stddev.cu b/cpp/test/stats/stddev.cu
index 998f7a88e7..cf57d3a923 100644
--- a/cpp/test/stats/stddev.cu
+++ b/cpp/test/stats/stddev.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,7 +15,7 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/matrix/math.cuh>
 #include <raft/random/rng.cuh>
@@ -23,6 +23,8 @@
 #include <raft/stats/stddev.cuh>
 #include <raft/util/cudart_utils.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/stats/sum.cu b/cpp/test/stats/sum.cu
index 290c53a563..5a549f8ba4 100644
--- a/cpp/test/stats/sum.cu
+++ b/cpp/test/stats/sum.cu
@@ -15,8 +15,8 @@
  */
 
 #include "../test_utils.cuh"
-#include <raft/core/resource/cuda_stream.hpp>
 
+#include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/eltwise.cuh>
 #include <raft/stats/sum.cuh>
diff --git a/cpp/test/stats/trustworthiness.cu b/cpp/test/stats/trustworthiness.cu
index 502a59f5c2..846c192022 100644
--- a/cpp/test/stats/trustworthiness.cu
+++ b/cpp/test/stats/trustworthiness.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,13 +15,15 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/distance/distance.cuh>
+#include <raft/stats/trustworthiness_score.cuh>
 #include <raft/util/cudart_utils.hpp>
 
-#include <raft/stats/trustworthiness_score.cuh>
+#include <gtest/gtest.h>
+
+#include <iostream>
 #include <vector>
 
 namespace raft {
diff --git a/cpp/test/stats/v_measure.cu b/cpp/test/stats/v_measure.cu
index 0cc164f277..c44df84ada 100644
--- a/cpp/test/stats/v_measure.cu
+++ b/cpp/test/stats/v_measure.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,13 +14,16 @@
  * limitations under the License.
  */
 #include "../test_utils.cuh"
-#include <algorithm>
-#include <gtest/gtest.h>
-#include <iostream>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/stats/homogeneity_score.cuh>
 #include <raft/stats/v_measure.cuh>
 #include <raft/util/cudart_utils.hpp>
+
+#include <gtest/gtest.h>
+
+#include <algorithm>
+#include <iostream>
 #include <random>
 
 namespace raft {
diff --git a/cpp/test/stats/weighted_mean.cu b/cpp/test/stats/weighted_mean.cu
index da1a825da1..407f3f14ea 100644
--- a/cpp/test/stats/weighted_mean.cu
+++ b/cpp/test/stats/weighted_mean.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,16 +15,20 @@
  */
 
 #include "../test_utils.cuh"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/random/rng.cuh>
 #include <raft/stats/weighted_mean.cuh>
 #include <raft/util/cuda_utils.cuh>
+
 #include <thrust/device_vector.h>
 #include <thrust/host_vector.h>
 
+#include <gtest/gtest.h>
+
+#include <cstdint>
+
 namespace raft {
 namespace stats {
 
diff --git a/cpp/test/test.cpp b/cpp/test/test.cpp
index 7477d7d0b5..b4727103f4 100644
--- a/cpp/test/test.cpp
+++ b/cpp/test/test.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,11 @@
  * limitations under the License.
  */
 
+#include <raft.hpp>
+
 #include <gtest/gtest.h>
+
 #include <iostream>
-#include <raft.hpp>
 
 namespace raft {
 
diff --git a/cpp/test/test_utils.cuh b/cpp/test/test_utils.cuh
index 1afa7acc83..810a0d7985 100644
--- a/cpp/test/test_utils.cuh
+++ b/cpp/test/test_utils.cuh
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,16 +18,19 @@
 
 #include "test_utils.h"
 
-#include <gtest/gtest.h>
-#include <iostream>
-#include <memory>
 #include <raft/random/rng.cuh>
 #include <raft/util/cuda_utils.cuh>
 #include <raft/util/cudart_utils.hpp>
+
 #include <rmm/exec_policy.hpp>
+
 #include <thrust/for_each.h>
 
+#include <gtest/gtest.h>
+
 #include <fstream>
+#include <iostream>
+#include <memory>
 #include <sstream>
 #include <stdexcept>
 #include <string>
diff --git a/cpp/test/test_utils.h b/cpp/test/test_utils.h
index 75590463b0..cf9a885cfb 100644
--- a/cpp/test/test_utils.h
+++ b/cpp/test/test_utils.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2018-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,12 +16,12 @@
 
 #pragma once
 
-#include <cmath>
-#include <iostream>
+#include <raft/core/kvp.hpp>
 
 #include <gtest/gtest.h>
 
-#include <raft/core/kvp.hpp>
+#include <cmath>
+#include <iostream>
 
 namespace raft {
 
diff --git a/cpp/test/util/cudart_utils.cpp b/cpp/test/util/cudart_utils.cpp
index 3fde494823..0ca05d5006 100644
--- a/cpp/test/util/cudart_utils.cpp
+++ b/cpp/test/util/cudart_utils.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -17,7 +17,6 @@
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/util/cudart_utils.hpp>
-#include <vector>
 
 #include <rmm/device_uvector.hpp>
 
@@ -25,6 +24,7 @@
 
 #include <iostream>
 #include <regex>
+#include <vector>
 
 namespace raft {
 
diff --git a/cpp/test/util/device_atomics.cu b/cpp/test/util/device_atomics.cu
index f1a97794ed..c5bb0ad3b6 100644
--- a/cpp/test/util/device_atomics.cu
+++ b/cpp/test/util/device_atomics.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2022-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2022-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,18 +14,21 @@
  * limitations under the License.
  */
 
+#include <raft/util/cudart_utils.hpp>
+#include <raft/util/device_atomics.cuh>
+
+#include <rmm/cuda_stream_pool.hpp>
+#include <rmm/device_scalar.hpp>
+#include <rmm/device_uvector.hpp>
+
+#include <gtest/gtest.h>
+
 #include <algorithm>
 #include <array>
 #include <cstddef>
-#include <gtest/gtest.h>
 #include <iostream>
 #include <memory>
 #include <numeric>
-#include <raft/util/cudart_utils.hpp>
-#include <raft/util/device_atomics.cuh>
-#include <rmm/cuda_stream_pool.hpp>
-#include <rmm/device_scalar.hpp>
-#include <rmm/device_uvector.hpp>
 
 namespace raft {
 
diff --git a/cpp/test/util/integer_utils.cpp b/cpp/test/util/integer_utils.cpp
index ed5dddf72d..7101fd2740 100644
--- a/cpp/test/util/integer_utils.cpp
+++ b/cpp/test/util/integer_utils.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,11 @@
  * limitations under the License.
  */
 
+#include <raft/util/integer_utils.hpp>
+
 #include <gtest/gtest.h>
+
 #include <iostream>
-#include <raft/util/integer_utils.hpp>
 
 namespace raft {
 
diff --git a/cpp/test/util/integer_utils.cu b/cpp/test/util/integer_utils.cu
index 1682bc9dc7..83f94d7fad 100644
--- a/cpp/test/util/integer_utils.cu
+++ b/cpp/test/util/integer_utils.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -15,11 +15,14 @@
  */
 
 #include "../test_utils.cuh"
-#include <gtest/gtest.h>
+
 #include <raft/core/resource/cuda_stream.hpp>
 #include <raft/util/integer_utils.hpp>
+
 #include <rmm/device_scalar.hpp>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 namespace util {
 
diff --git a/cpp/test/util/memory_type_dispatcher.cu b/cpp/test/util/memory_type_dispatcher.cu
index 5e24ff5719..5b8d36af73 100644
--- a/cpp/test/util/memory_type_dispatcher.cu
+++ b/cpp/test/util/memory_type_dispatcher.cu
@@ -15,8 +15,7 @@
  */
 
 #include "../test_utils.h"
-#include <cstdint>
-#include <gtest/gtest.h>
+
 #include <raft/core/device_mdarray.hpp>
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/device_resources.hpp>
@@ -28,6 +27,10 @@
 #include <raft/core/pinned_mdarray.hpp>
 #include <raft/core/pinned_mdspan.hpp>
 #include <raft/util/memory_type_dispatcher.cuh>
+
+#include <gtest/gtest.h>
+
+#include <cstdint>
 #include <utility>
 #include <variant>
 
diff --git a/cpp/test/util/pow2_utils.cu b/cpp/test/util/pow2_utils.cu
index e29e4eeb9c..f618852596 100644
--- a/cpp/test/util/pow2_utils.cu
+++ b/cpp/test/util/pow2_utils.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2023, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2024, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,9 +14,10 @@
  * limitations under the License.
  */
 
-#include <gtest/gtest.h>
 #include <raft/util/pow2_utils.cuh>
 
+#include <gtest/gtest.h>
+
 namespace raft {
 
 template <auto Val, typename TargetT>
diff --git a/dependencies.yaml b/dependencies.yaml
index a451b6836f..72aa3427d1 100644
--- a/dependencies.yaml
+++ b/dependencies.yaml
@@ -392,13 +392,17 @@ dependencies:
             packages:
               - python=3.10
           - matrix:
+              py: "3.11"
             packages:
-              - python>=3.9,<3.11
+              - python=3.11
+          - matrix:
+            packages:
+              - python>=3.9,<3.12
   run_pylibraft:
     common:
       - output_types: [conda, pyproject]
         packages:
-          - &numpy numpy>=1.21
+          - &numpy numpy>=1.23
       - output_types: [conda]
         packages:
           - *rmm_conda
diff --git a/python/pylibraft/pylibraft/test/test_ivf_flat.py b/python/pylibraft/pylibraft/test/test_ivf_flat.py
index 23140073f1..2e38dab7bc 100644
--- a/python/pylibraft/pylibraft/test/test_ivf_flat.py
+++ b/python/pylibraft/pylibraft/test/test_ivf_flat.py
@@ -1,4 +1,4 @@
-# Copyright (c) 2023, NVIDIA CORPORATION.
+# Copyright (c) 2023-2024, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -297,6 +297,14 @@ def test_ivf_flat_params(params):
             "k": 129,
             "n_probes": 100,
         },
+        {
+            "k": 257,
+            "n_probes": 100,
+        },
+        {
+            "k": 4096,
+            "n_probes": 100,
+        },
     ],
 )
 def test_ivf_pq_search_params(params):
diff --git a/python/pylibraft/pyproject.toml b/python/pylibraft/pyproject.toml
index 5e9701a47a..6468220330 100644
--- a/python/pylibraft/pyproject.toml
+++ b/python/pylibraft/pyproject.toml
@@ -36,7 +36,7 @@ license = { text = "Apache 2.0" }
 requires-python = ">=3.9"
 dependencies = [
     "cuda-python>=11.7.1,<12.0a0",
-    "numpy>=1.21",
+    "numpy>=1.23",
     "rmm==24.4.*",
 ] # This list was generated by `rapids-dependency-file-generator`. To make changes, edit ../../dependencies.yaml and run `rapids-dependency-file-generator`.
 classifiers = [
@@ -44,6 +44,7 @@ classifiers = [
     "Programming Language :: Python",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
 ]
 
 [project.optional-dependencies]
diff --git a/python/raft-dask/pyproject.toml b/python/raft-dask/pyproject.toml
index bd256e8e92..b869290d5c 100644
--- a/python/raft-dask/pyproject.toml
+++ b/python/raft-dask/pyproject.toml
@@ -36,7 +36,7 @@ dependencies = [
     "dask-cuda==24.4.*",
     "joblib>=0.11",
     "numba>=0.57",
-    "numpy>=1.21",
+    "numpy>=1.23",
     "pylibraft==24.4.*",
     "rapids-dask-dependency==24.4.*",
     "ucx-py==0.37.*",
@@ -46,6 +46,7 @@ classifiers = [
     "Programming Language :: Python",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
 ]
 
 [project.optional-dependencies]