Comment by nextaccountic

I don't want a CUDA rival. I want to get the entire pile of CUDA code that is already written and run it on AMD GPUs without any kind of tweak or rewrite, and have it just work every time

Compatibility with existing code is very important. People can't afford to rewrite their stuff just to support AMD, and thus they don't

AMD is kind of trying to do this with rocm and HIP, but whatever they are doing it's not enough