jcinPDX@pytorch 2.4 upstream now includes a prototype feature supporting Intel GPUs through source build using #SYCL and #oneDNN as well as a backend integrated to inductor on top of Triton - enabling a path for millions and millions of GPUs through #oneAPI for #AI.
Lots of important milestones to make this happen - including support for #UXL Foundation open AI technologies. Just a prototype, but a big step forward... thanks to all in the PyTorch community. Feedback welcome!
PyTorch 2.4 Release Blog
We are excited to announce the release of PyTorch® 2.4 (release note)! PyTorch 2.4 adds support for the latest version of Python (3.12) for torch.compile. AOTInductor freezing gives developers running AOTInductor more performance-based optimizations by allowing the serialization of MKLDNN weights. As well, a new default TCPStore server backend utilizing libuv has been introduced which should significantly reduce initialization times for users running large-scale jobs. Finally, a new Python Custom Operator API makes it easier than before to integrate custom kernels into PyTorch, especially for torch.compile.
Phoronix
