diff --git a/examples/pytorch_optimization.py b/examples/pytorch_optimization.py
index f731cc4b..658cd485 100644
--- a/examples/pytorch_optimization.py
+++ b/examples/pytorch_optimization.py
@@ -16,7 +16,7 @@
 import os
 import torch
 import matplotlib
-matplotlib.use('agg')
+matplotlib.use('agg') # Make matplotlib more robust when interface plotting is impossible.
 import matplotlib.pyplot as plt
 import argparse
 
diff --git a/setup.py b/setup.py
index 6857b9bc..f4e3bffb 100644
--- a/setup.py
+++ b/setup.py
@@ -22,6 +22,7 @@
 import sys
 import textwrap
 import traceback
+from typing import List
 
 
 from distutils.errors import CompileError, DistutilsError, \
@@ -437,21 +438,26 @@ def is_torch_cuda(build_ext, include_dirs, extra_compile_args):
         print('INFO: Above error indicates that this PyTorch installation does not support CUDA.')
         return False
 
-def build_nvcc_extra_objects(cxx11_abi: bool):
+def get_nvcc_cmd() -> str:
+    from shutil import which
+    nvcc_cmd = which('nvcc')
+    if nvcc_cmd is None:
+        raise DistutilsPlatformError('Unable to find NVCC compiler')
+    return nvcc_cmd
+
+def build_nvcc_extra_objects(nvcc_cmd: str, cxx11_abi: bool) -> List[str]:
     # nvcc --compiler-options '-fPIC -D_GLIBCXX_USE_CXX11_ABI=0' -rdc=true -c cuda_kernels.cu
     # nvcc --compiler-options '-fPIC -D_GLIBCXX_USE_CXX11_ABI=0' -dlink -o cuda_kernels_link.o \
     #      cuda_kernels.o -lcudart
-    nvcc = 'nvcc'
-    common_flags = f"'-fPIC -D_GLIBCXX_USE_CXX11_ABI={int(cxx11_abi)}'"
-    nvcc_with_flags = f"{nvcc} --compiler-options {common_flags}"
+    nvcc_flags = f"--compiler-options '-fPIC -D_GLIBCXX_USE_CXX11_ABI={int(cxx11_abi)}'"
 
     extra_object_dir = 'bluefog/common/cuda/'
     source = extra_object_dir+'cuda_kernels.cu'
     object_file = extra_object_dir+'cuda_kernels.o'
     object_link = extra_object_dir+'cuda_kernels_link.o'
 
-    command_object = f"{nvcc_with_flags} -rdc=true -c {source} -o {object_file}"
-    command_link   = f"{nvcc_with_flags} -dlink {object_file} -lcudart -o {object_link}"
+    command_object = f"{nvcc_cmd} {nvcc_flags} -rdc=true -c {source} -o {object_file}"
+    command_link   = f"{nvcc_cmd} {nvcc_flags} -dlink {object_file} -lcudart -o {object_link}"
 
     subprocess.check_call([command_object], shell=True)
     subprocess.check_call([command_link], shell=True)
@@ -469,7 +475,8 @@ def build_torch_extension(build_ext, global_options, torch_version):
     if have_cuda:
         cuda_include_dirs, cuda_lib_dirs = get_cuda_dirs(
             build_ext, options['COMPILE_FLAGS'])
-        cuda_extra_objects = build_nvcc_extra_objects(is_cxx11_abi)
+        nvcc_cmd = get_nvcc_cmd()
+        cuda_extra_objects = build_nvcc_extra_objects(nvcc_cmd, is_cxx11_abi)
         options['EXTRA_OBJECTS'] += cuda_extra_objects
 
         options['INCLUDES'] += cuda_include_dirs
@@ -515,7 +522,6 @@ def build_torch_extension(build_ext, global_options, torch_version):
     else:
         # CUDAExtension fails with `ld: library not found for -lcudart` if CUDA is not present
         from torch.utils.cpp_extension import CppExtension as TorchExtension
-    bluefog_tensorflow_mpi_lib.extra_objects = options['EXTRA_OBJECTS']
 
     ext = TorchExtension(bluefog_torch_mpi_lib.name,
                          define_macros=updated_macros,