OpenCL works on both AMD and Nvidia GPUs with mostly the same source code. By supporting at-runtime compilation it allows a lot of code particularization/instantiation before compilation, which reduces the power (cost) of the generated code. In general OpenCL is close enough to the HW and the generated code is improving over time (LLVM).
Motivation: a long time ago I had an AMD GPU and no way to run an LL test on it, so I decided to write my own. And I was hooked by the power of the GPU and the quest for ever more efficient, faster implem.
Comments
OpenCL works on both AMD and Nvidia GPUs with mostly the same source code. By supporting at-runtime compilation it allows a lot of code particularization/instantiation before compilation, which reduces the power (cost) of the generated code. In general OpenCL is close enough to the HW and the generated code is improving over time (LLVM).
Motivation: a long time ago I had an AMD GPU and no way to run an LL test on it, so I decided to write my own. And I was hooked by the power of the GPU and the quest for ever more efficient, faster implem.