Freestanding c++ with compiler intrinsics is a nicer alternative. You can do things like take the address of a function. Use an interface over memory allocatio

Freestanding c++ with compiler intrinsics is a nicer alternative. You can do things like take the address of a function. Use an interface over memory allocation/queue launch with implementations in cuda, hsa, opencl whatever. All the rest of the GPU side stuff is syntax sugar/salt over slightly weird semantics, totally possible to opt out of all of that.