[rnrn]

or look at using standard C++ with an implementation that uses the GPU: https://docs.nvidia.com/hpc-sdk/compilers/c++-parallel-algor...