Great - a few references for some background:
- https://github.com/gmarkall/extending-numba-cuda/blob/main/Extending%20Numba’s%20CUDA%20Target.ipynb
- The Life of a Numba Kernel: A Compilation Pipeline Taking User Defined Functions in Python to CUDA Kernels | by Graham Markall | RAPIDS AI | Medium
- Contributing to Numba with no compiler or LLVM experience
If you want to take a look at these to get a feel for what’s generally required and see whether it looks like something that would fit your interest, then I can provide a bit more guidance (having to be brief here now as it’s getting a little late) - do let me know what you think!