pybind11-cuda

Starting point for GPU accelerated python libraries

Adapted from original work from https://github.com/PWhiddy/pybind11-cuda

Present work uses modern CMake/Cuda approach

Prerequisites

Cuda

Python 3.6 or greater

Cmake >= 3.12 (for CUDA support and the new FindPython3 module)

Build instructions

cmake

If you use cmake version >= 3.18, you can use variable CMAKE_CUDA_ARCHITECTURES instead of CUDAFLAGS:

mkdir build; cd build
# provide a default cuda hardware architecture to build for
cmake -DCMAKE_CUDA_ARCHITECTURES="75" -DPython3_EXECUTABLE=`which python` ..
make

Please note that specifiying Python3_EXECUTABLE is not required, but recommended if you have multiple python executable on your system (e.g. one from OS, another from conda, etc...); this way you can control which python installation will be used.

If you have an older version cmake, you can pass nvcc flags to cmake using env variable CUDAFLAGS

# change CUDAFLAGS according to your target GPU architecture
mkdir build; cd build
# provide a default cuda hardware architecture to build for
export CUDAFLAGS="-arch=sm_75"
cmake -DPython3_EXECUTABLE=`which python` ..
make

test

Test it with

cd src
python3 test_mul.py

gpu_library.so and test_mul.py must be in the same folder. Alternatively you can path to gpu_library.so to your PYTHONPATH env variable.

Features demonstrated

• Compiles out of the box with cmake
• Numpy integration
• C++ Templating for composable kernels with generic data types

Originally based on https://github.com/torstem/demo-cuda-pybind11