Something that is hinted at, but not spelled out loud in our posts is that AMD actively upstreams and maintains a LLVM back-end for their GPUs, so it really is

Something that is hinted at, but not spelled out loud in our posts is that AMD actively upstreams and maintains a LLVM back-end for their GPUs, so it really is a matter of switching the binary target for the generated code, at least in theory :)