Small example of how to compile & use cuda kernels with Rust using both nalgebra and ndarray.
'src/lib.rs' shows how to use the kernel in 'cuda/matrix.cu'.
To compile the CUDA code and run the test:
$ cargo test
I've added CUDA-specific tests/benchmarks using Google Test and Google Benchmark so the kernels can be tested outside of the Rust code. Since they use git submodules you may want to clone this repository with:
$ git clone --recursive git@github.com:PhDP/rust-cuda-template.git
You are free to clone this repository and adopt any license you wish to adopt for your code.