Intel: AMD joins UXL Foundation as steering member to advance oneDNN PyTorch work; AMD and Intel have aligned on oneDNN as the default open-source library for x86 acceleration within PyTorch