AMD and Intel Collaborate to Simplify AI Development on X86 CPUs Using PyTorch and OneDNN
AMD and Intel have partnered to simplify AI development on X86 CPUs by aligning on OneDNN as the default open-source library for X86 acceleration within PyTorch. AMD has also joined the UXL Foundation as a steering member to advance OneDNN PyTorch work, enhancing collaboration in AI software development for X86 architectures.