[almostgotcaught]

> its either CUDA or HIP, which differ widely in parts of their compilation stack.

This is an ironic comment - OpenCL uses the same compiler as CUDA on NVIDIA and HIP on AMD.

[JonChesterfield]

Sort of. Same compiler backend, mostly, but the set of intrinsics and semantic rules are different.

[almostgotcaught]

i have no idea what your point is - same compiler, different frontend, yes that's literally what i said.