diff --git a/cpp/benchmarks/CMakeLists.txt b/cpp/benchmarks/CMakeLists.txt
index 10f645dfec0..49504e53424 100644
--- a/cpp/benchmarks/CMakeLists.txt
+++ b/cpp/benchmarks/CMakeLists.txt
@@ -163,8 +163,10 @@ ConfigureNVBench(
 
 # ##################################################################################################
 # * join benchmark --------------------------------------------------------------------------------
-ConfigureBench(JOIN_BENCH join/left_join.cu join/conditional_join.cu)
-ConfigureNVBench(JOIN_NVBENCH join/join.cu join/mixed_join.cu join/distinct_join.cu)
+ConfigureNVBench(
+  JOIN_NVBENCH join/left_join.cu join/conditional_join.cu join/join.cu join/mixed_join.cu
+  join/distinct_join.cu
+)
 
 # ##################################################################################################
 # * iterator benchmark ----------------------------------------------------------------------------
diff --git a/cpp/benchmarks/join/conditional_join.cu b/cpp/benchmarks/join/conditional_join.cu
index d95fc0a5b59..e332d09d31b 100644
--- a/cpp/benchmarks/join/conditional_join.cu
+++ b/cpp/benchmarks/join/conditional_join.cu
@@ -14,250 +14,44 @@
  * limitations under the License.
  */
 
-#include <benchmarks/join/join_common.hpp>
-
-template <typename Key>
-class ConditionalJoin : public cudf::benchmark {};
-
-// For compatibility with the shared logic for equality (hash) joins, all of
-// the join lambdas defined by these macros accept a null_equality parameter
-// but ignore it (don't forward it to the underlying join implementation)
-// because conditional joins do not use this parameter.
-#define CONDITIONAL_INNER_JOIN_BENCHMARK_DEFINE(name, Key, Nullable) \
-  BENCHMARK_TEMPLATE_DEFINE_F(ConditionalJoin, name, Key)            \
-  (::benchmark::State & st)                                          \
-  {                                                                  \
-    auto join = [](cudf::table_view const& left,                     \
-                   cudf::table_view const& right,                    \
-                   cudf::ast::operation binary_pred,                 \
-                   cudf::null_equality compare_nulls) {              \
-      return cudf::conditional_inner_join(left, right, binary_pred); \
-    };                                                               \
-    BM_join<Key, Nullable, join_t::CONDITIONAL>(st, join);           \
-  }
-
-CONDITIONAL_INNER_JOIN_BENCHMARK_DEFINE(conditional_inner_join_32bit, int32_t, false);
-CONDITIONAL_INNER_JOIN_BENCHMARK_DEFINE(conditional_inner_join_64bit, int64_t, false);
-CONDITIONAL_INNER_JOIN_BENCHMARK_DEFINE(conditional_inner_join_32bit_nulls, int32_t, true);
-CONDITIONAL_INNER_JOIN_BENCHMARK_DEFINE(conditional_inner_join_64bit_nulls, int64_t, true);
-
-#define CONDITIONAL_LEFT_JOIN_BENCHMARK_DEFINE(name, Key, Nullable) \
-  BENCHMARK_TEMPLATE_DEFINE_F(ConditionalJoin, name, Key)           \
-  (::benchmark::State & st)                                         \
-  {                                                                 \
-    auto join = [](cudf::table_view const& left,                    \
-                   cudf::table_view const& right,                   \
-                   cudf::ast::operation binary_pred,                \
-                   cudf::null_equality compare_nulls) {             \
-      return cudf::conditional_left_join(left, right, binary_pred); \
-    };                                                              \
-    BM_join<Key, Nullable, join_t::CONDITIONAL>(st, join);          \
-  }
-
-CONDITIONAL_LEFT_JOIN_BENCHMARK_DEFINE(conditional_left_join_32bit, int32_t, false);
-CONDITIONAL_LEFT_JOIN_BENCHMARK_DEFINE(conditional_left_join_64bit, int64_t, false);
-CONDITIONAL_LEFT_JOIN_BENCHMARK_DEFINE(conditional_left_join_32bit_nulls, int32_t, true);
-CONDITIONAL_LEFT_JOIN_BENCHMARK_DEFINE(conditional_left_join_64bit_nulls, int64_t, true);
-
-#define CONDITIONAL_FULL_JOIN_BENCHMARK_DEFINE(name, Key, Nullable) \
-  BENCHMARK_TEMPLATE_DEFINE_F(ConditionalJoin, name, Key)           \
-  (::benchmark::State & st)                                         \
-  {                                                                 \
-    auto join = [](cudf::table_view const& left,                    \
-                   cudf::table_view const& right,                   \
-                   cudf::ast::operation binary_pred,                \
-                   cudf::null_equality compare_nulls) {             \
-      return cudf::conditional_full_join(left, right, binary_pred); \
-    };                                                              \
-    BM_join<Key, Nullable, join_t::CONDITIONAL>(st, join);          \
-  }
-
-CONDITIONAL_FULL_JOIN_BENCHMARK_DEFINE(conditional_full_join_32bit, int32_t, false);
-CONDITIONAL_FULL_JOIN_BENCHMARK_DEFINE(conditional_full_join_64bit, int64_t, false);
-CONDITIONAL_FULL_JOIN_BENCHMARK_DEFINE(conditional_full_join_32bit_nulls, int32_t, true);
-CONDITIONAL_FULL_JOIN_BENCHMARK_DEFINE(conditional_full_join_64bit_nulls, int64_t, true);
-
-#define CONDITIONAL_LEFT_ANTI_JOIN_BENCHMARK_DEFINE(name, Key, Nullable) \
-  BENCHMARK_TEMPLATE_DEFINE_F(ConditionalJoin, name, Key)                \
-  (::benchmark::State & st)                                              \
-  {                                                                      \
-    auto join = [](cudf::table_view const& left,                         \
-                   cudf::table_view const& right,                        \
-                   cudf::ast::operation binary_pred,                     \
-                   cudf::null_equality compare_nulls) {                  \
-      return cudf::conditional_left_anti_join(left, right, binary_pred); \
-    };                                                                   \
-    BM_join<Key, Nullable, join_t::CONDITIONAL>(st, join);               \
-  }
-
-CONDITIONAL_LEFT_ANTI_JOIN_BENCHMARK_DEFINE(conditional_left_anti_join_32bit, int32_t, false);
-CONDITIONAL_LEFT_ANTI_JOIN_BENCHMARK_DEFINE(conditional_left_anti_join_64bit, int64_t, false);
-CONDITIONAL_LEFT_ANTI_JOIN_BENCHMARK_DEFINE(conditional_left_anti_join_32bit_nulls, int32_t, true);
-CONDITIONAL_LEFT_ANTI_JOIN_BENCHMARK_DEFINE(conditional_left_anti_join_64bit_nulls, int64_t, true);
-
-#define CONDITIONAL_LEFT_SEMI_JOIN_BENCHMARK_DEFINE(name, Key, Nullable) \
-  BENCHMARK_TEMPLATE_DEFINE_F(ConditionalJoin, name, Key)                \
-  (::benchmark::State & st)                                              \
-  {                                                                      \
-    auto join = [](cudf::table_view const& left,                         \
-                   cudf::table_view const& right,                        \
-                   cudf::ast::operation binary_pred,                     \
-                   cudf::null_equality compare_nulls) {                  \
-      return cudf::conditional_left_semi_join(left, right, binary_pred); \
-    };                                                                   \
-    BM_join<Key, Nullable, join_t::CONDITIONAL>(st, join);               \
-  }
-
-CONDITIONAL_LEFT_SEMI_JOIN_BENCHMARK_DEFINE(conditional_left_semi_join_32bit, int32_t, false);
-CONDITIONAL_LEFT_SEMI_JOIN_BENCHMARK_DEFINE(conditional_left_semi_join_64bit, int64_t, false);
-CONDITIONAL_LEFT_SEMI_JOIN_BENCHMARK_DEFINE(conditional_left_semi_join_32bit_nulls, int32_t, true);
-CONDITIONAL_LEFT_SEMI_JOIN_BENCHMARK_DEFINE(conditional_left_semi_join_64bit_nulls, int64_t, true);
-
-// inner join -----------------------------------------------------------------------
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_inner_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({400'000, 100'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_inner_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({400'000, 100'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_inner_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({400'000, 100'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_inner_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({400'000, 100'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-// left join -----------------------------------------------------------------------
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-// full join -----------------------------------------------------------------------
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_full_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_full_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_full_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_full_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-// left anti-join -------------------------------------------------------------
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_anti_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_anti_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_anti_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_anti_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-// left semi-join -------------------------------------------------------------
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_semi_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_semi_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_semi_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(ConditionalJoin, conditional_left_semi_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->UseManualTime();
+#include "join_common.hpp"
+
+template <typename Key, bool Nullable>
+void nvbench_conditional_inner_join(nvbench::state& state,
+                                    nvbench::type_list<Key, nvbench::enum_type<Nullable>>)
+{
+  auto join = [](cudf::table_view const& left,
+                 cudf::table_view const& right,
+                 cudf::ast::operation binary_pred,
+                 cudf::null_equality compare_nulls) {
+    return cudf::conditional_inner_join(left, right, binary_pred);
+  };
+  BM_join<Key, Nullable, join_t::CONDITIONAL>(state, join);
+}
+
+template <typename Key, bool Nullable>
+void nvbench_conditional_left_join(nvbench::state& state,
+                                   nvbench::type_list<Key, nvbench::enum_type<Nullable>>)
+{
+  auto join = [](cudf::table_view const& left,
+                 cudf::table_view const& right,
+                 cudf::ast::operation binary_pred,
+                 cudf::null_equality compare_nulls) {
+    return cudf::conditional_left_join(left, right, binary_pred);
+  };
+  BM_join<Key, Nullable, join_t::CONDITIONAL>(state, join);
+}
+
+NVBENCH_BENCH_TYPES(nvbench_conditional_inner_join,
+                    NVBENCH_TYPE_AXES(JOIN_KEY_TYPE_RANGE, JOIN_NULLABLE_RANGE))
+  .set_name("conditional_inner_join")
+  .set_type_axes_names({"Key", "Nullable"})
+  .add_int64_axis("left_size", JOIN_SIZE_RANGE)
+  .add_int64_axis("right_size", JOIN_SIZE_RANGE);
+
+NVBENCH_BENCH_TYPES(nvbench_conditional_left_join,
+                    NVBENCH_TYPE_AXES(JOIN_KEY_TYPE_RANGE, JOIN_NULLABLE_RANGE))
+  .set_name("conditional_left_join")
+  .set_type_axes_names({"Key", "Nullable"})
+  .add_int64_axis("left_size", JOIN_SIZE_RANGE)
+  .add_int64_axis("right_size", JOIN_SIZE_RANGE);
diff --git a/cpp/benchmarks/join/join_common.hpp b/cpp/benchmarks/join/join_common.hpp
index e6792b9dbfb..3d9d9c57548 100644
--- a/cpp/benchmarks/join/join_common.hpp
+++ b/cpp/benchmarks/join/join_common.hpp
@@ -19,7 +19,6 @@
 #include "generate_input_tables.cuh"
 
 #include <benchmarks/fixture/benchmark_fixture.hpp>
-#include <benchmarks/synchronization/synchronization.hpp>
 
 #include <cudf/ast/expressions.hpp>
 #include <cudf/column/column_factories.hpp>
@@ -67,28 +66,12 @@ template <typename Key,
           typename Join>
 void BM_join(state_type& state, Join JoinFunc)
 {
-  auto const right_size = [&]() {
-    if constexpr (std::is_same_v<state_type, benchmark::State>) {
-      return static_cast<cudf::size_type>(state.range(0));
-    }
-    if constexpr (std::is_same_v<state_type, nvbench::state>) {
-      return static_cast<cudf::size_type>(state.get_int64("right_size"));
-    }
-  }();
-  auto const left_size = [&]() {
-    if constexpr (std::is_same_v<state_type, benchmark::State>) {
-      return static_cast<cudf::size_type>(state.range(1));
-    }
-    if constexpr (std::is_same_v<state_type, nvbench::state>) {
-      return static_cast<cudf::size_type>(state.get_int64("left_size"));
-    }
-  }();
+  auto const right_size = static_cast<cudf::size_type>(state.get_int64("right_size"));
+  auto const left_size  = static_cast<cudf::size_type>(state.get_int64("left_size"));
 
-  if constexpr (std::is_same_v<state_type, nvbench::state>) {
-    if (right_size > left_size) {
-      state.skip("Skip large right table");
-      return;
-    }
+  if (right_size > left_size) {
+    state.skip("Skip large right table");
+    return;
   }
 
   double const selectivity = 0.3;
@@ -165,57 +148,37 @@ void BM_join(state_type& state, Join JoinFunc)
 
   // Setup join parameters and result table
   [[maybe_unused]] std::vector<cudf::size_type> columns_to_join = {0};
-
-  // Benchmark the inner join operation
-  if constexpr (std::is_same_v<state_type, benchmark::State> and
-                (join_type != join_t::CONDITIONAL)) {
-    for (auto _ : state) {
-      cuda_event_timer raii(state, true, cudf::get_default_stream());
-
-      auto result = JoinFunc(left_table.select(columns_to_join),
-                             right_table.select(columns_to_join),
-                             cudf::null_equality::UNEQUAL);
-    }
-  }
-  if constexpr (std::is_same_v<state_type, nvbench::state> and (join_type != join_t::CONDITIONAL)) {
-    state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
-    if constexpr (join_type == join_t::MIXED) {
-      auto const col_ref_left_0 = cudf::ast::column_reference(0);
-      auto const col_ref_right_0 =
-        cudf::ast::column_reference(0, cudf::ast::table_reference::RIGHT);
-      auto left_zero_eq_right_zero =
-        cudf::ast::operation(cudf::ast::ast_operator::EQUAL, col_ref_left_0, col_ref_right_0);
-      state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
-        auto result = JoinFunc(left_table.select(columns_to_join),
-                               right_table.select(columns_to_join),
-                               left_table.select({1}),
-                               right_table.select({1}),
-                               left_zero_eq_right_zero,
-                               cudf::null_equality::UNEQUAL);
-      });
-    }
-    if constexpr (join_type == join_t::HASH) {
-      state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
-        auto result = JoinFunc(left_table.select(columns_to_join),
-                               right_table.select(columns_to_join),
-                               cudf::null_equality::UNEQUAL);
-      });
-    }
-  }
-
-  // Benchmark conditional join
-  if constexpr (std::is_same_v<state_type, benchmark::State> and join_type == join_t::CONDITIONAL) {
-    // Common column references.
+  state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
+  if constexpr (join_type == join_t::CONDITIONAL) {
     auto const col_ref_left_0  = cudf::ast::column_reference(0);
     auto const col_ref_right_0 = cudf::ast::column_reference(0, cudf::ast::table_reference::RIGHT);
     auto left_zero_eq_right_zero =
       cudf::ast::operation(cudf::ast::ast_operator::EQUAL, col_ref_left_0, col_ref_right_0);
-
-    for (auto _ : state) {
-      cuda_event_timer raii(state, true, cudf::get_default_stream());
-
+    state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
       auto result =
         JoinFunc(left_table, right_table, left_zero_eq_right_zero, cudf::null_equality::UNEQUAL);
-    }
+      ;
+    });
+  }
+  if constexpr (join_type == join_t::MIXED) {
+    auto const col_ref_left_0  = cudf::ast::column_reference(0);
+    auto const col_ref_right_0 = cudf::ast::column_reference(0, cudf::ast::table_reference::RIGHT);
+    auto left_zero_eq_right_zero =
+      cudf::ast::operation(cudf::ast::ast_operator::EQUAL, col_ref_left_0, col_ref_right_0);
+    state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
+      auto result = JoinFunc(left_table.select(columns_to_join),
+                             right_table.select(columns_to_join),
+                             left_table.select({1}),
+                             right_table.select({1}),
+                             left_zero_eq_right_zero,
+                             cudf::null_equality::UNEQUAL);
+    });
+  }
+  if constexpr (join_type == join_t::HASH) {
+    state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
+      auto result = JoinFunc(left_table.select(columns_to_join),
+                             right_table.select(columns_to_join),
+                             cudf::null_equality::UNEQUAL);
+    });
   }
 }
diff --git a/cpp/benchmarks/join/left_join.cu b/cpp/benchmarks/join/left_join.cu
index 3e398e721fa..92123ce1621 100644
--- a/cpp/benchmarks/join/left_join.cu
+++ b/cpp/benchmarks/join/left_join.cu
@@ -14,115 +14,43 @@
  * limitations under the License.
  */
 
-#include <benchmarks/join/join_common.hpp>
-
-template <typename Key>
-class Join : public cudf::benchmark {};
-
-#define LEFT_ANTI_JOIN_BENCHMARK_DEFINE(name, Key, Nullable)   \
-  BENCHMARK_TEMPLATE_DEFINE_F(Join, name, Key)                 \
-  (::benchmark::State & st)                                    \
-  {                                                            \
-    auto join = [](cudf::table_view const& left,               \
-                   cudf::table_view const& right,              \
-                   cudf::null_equality compare_nulls) {        \
-      return cudf::left_anti_join(left, right, compare_nulls); \
-    };                                                         \
-    BM_join<Key, Nullable>(st, join);                          \
-  }
-
-LEFT_ANTI_JOIN_BENCHMARK_DEFINE(left_anti_join_32bit, int32_t, false);
-LEFT_ANTI_JOIN_BENCHMARK_DEFINE(left_anti_join_64bit, int64_t, false);
-LEFT_ANTI_JOIN_BENCHMARK_DEFINE(left_anti_join_32bit_nulls, int32_t, true);
-LEFT_ANTI_JOIN_BENCHMARK_DEFINE(left_anti_join_64bit_nulls, int64_t, true);
-
-#define LEFT_SEMI_JOIN_BENCHMARK_DEFINE(name, Key, Nullable)   \
-  BENCHMARK_TEMPLATE_DEFINE_F(Join, name, Key)                 \
-  (::benchmark::State & st)                                    \
-  {                                                            \
-    auto join = [](cudf::table_view const& left,               \
-                   cudf::table_view const& right,              \
-                   cudf::null_equality compare_nulls) {        \
-      return cudf::left_semi_join(left, right, compare_nulls); \
-    };                                                         \
-    BM_join<Key, Nullable>(st, join);                          \
-  }
-
-LEFT_SEMI_JOIN_BENCHMARK_DEFINE(left_semi_join_32bit, int32_t, false);
-LEFT_SEMI_JOIN_BENCHMARK_DEFINE(left_semi_join_64bit, int64_t, false);
-LEFT_SEMI_JOIN_BENCHMARK_DEFINE(left_semi_join_32bit_nulls, int32_t, true);
-LEFT_SEMI_JOIN_BENCHMARK_DEFINE(left_semi_join_64bit_nulls, int64_t, true);
-
-// left anti-join -------------------------------------------------------------
-BENCHMARK_REGISTER_F(Join, left_anti_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->Args({10'000'000, 10'000'000})
-  ->Args({10'000'000, 40'000'000})
-  ->Args({10'000'000, 100'000'000})
-  ->Args({100'000'000, 100'000'000})
-  ->Args({80'000'000, 240'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_anti_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({50'000'000, 50'000'000})
-  ->Args({40'000'000, 120'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_anti_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->Args({10'000'000, 10'000'000})
-  ->Args({10'000'000, 40'000'000})
-  ->Args({10'000'000, 100'000'000})
-  ->Args({100'000'000, 100'000'000})
-  ->Args({80'000'000, 240'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_anti_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({50'000'000, 50'000'000})
-  ->Args({40'000'000, 120'000'000})
-  ->UseManualTime();
-
-// left semi-join -------------------------------------------------------------
-BENCHMARK_REGISTER_F(Join, left_semi_join_32bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->Args({10'000'000, 10'000'000})
-  ->Args({10'000'000, 40'000'000})
-  ->Args({10'000'000, 100'000'000})
-  ->Args({100'000'000, 100'000'000})
-  ->Args({80'000'000, 240'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_semi_join_64bit)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({50'000'000, 50'000'000})
-  ->Args({40'000'000, 120'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_semi_join_32bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({100'000, 100'000})
-  ->Args({100'000, 400'000})
-  ->Args({100'000, 1'000'000})
-  ->Args({10'000'000, 10'000'000})
-  ->Args({10'000'000, 40'000'000})
-  ->Args({10'000'000, 100'000'000})
-  ->Args({100'000'000, 100'000'000})
-  ->Args({80'000'000, 240'000'000})
-  ->UseManualTime();
-
-BENCHMARK_REGISTER_F(Join, left_semi_join_64bit_nulls)
-  ->Unit(benchmark::kMillisecond)
-  ->Args({50'000'000, 50'000'000})
-  ->Args({40'000'000, 120'000'000})
-  ->UseManualTime();
+#include "join_common.hpp"
+
+template <typename Key, bool Nullable>
+void nvbench_left_anti_join(nvbench::state& state,
+                            nvbench::type_list<Key, nvbench::enum_type<Nullable>>)
+{
+  auto join = [](cudf::table_view const& left,
+                 cudf::table_view const& right,
+                 cudf::null_equality compare_nulls) {
+    return cudf::left_anti_join(left, right, compare_nulls);
+  };
+
+  BM_join<Key, Nullable>(state, join);
+}
+
+template <typename Key, bool Nullable>
+void nvbench_left_semi_join(nvbench::state& state,
+                            nvbench::type_list<Key, nvbench::enum_type<Nullable>>)
+{
+  auto join = [](cudf::table_view const& left,
+                 cudf::table_view const& right,
+                 cudf::null_equality compare_nulls) {
+    return cudf::left_semi_join(left, right, compare_nulls);
+  };
+  BM_join<Key, Nullable>(state, join);
+}
+
+NVBENCH_BENCH_TYPES(nvbench_left_anti_join,
+                    NVBENCH_TYPE_AXES(JOIN_KEY_TYPE_RANGE, JOIN_NULLABLE_RANGE))
+  .set_name("left_anti_join")
+  .set_type_axes_names({"Key", "Nullable"})
+  .add_int64_axis("left_size", JOIN_SIZE_RANGE)
+  .add_int64_axis("right_size", JOIN_SIZE_RANGE);
+
+NVBENCH_BENCH_TYPES(nvbench_left_semi_join,
+                    NVBENCH_TYPE_AXES(JOIN_KEY_TYPE_RANGE, JOIN_NULLABLE_RANGE))
+  .set_name("left_semi_join")
+  .set_type_axes_names({"Key", "Nullable"})
+  .add_int64_axis("left_size", JOIN_SIZE_RANGE)
+  .add_int64_axis("right_size", JOIN_SIZE_RANGE);