PyBindToGPUs. Parallel Computing starter project to build GPU & CPU kernels in CUDA & C++ and call them from Python without a single line of CMake using PyBind11