Intel and AMD Team Up to Simplify x86 CPU AI Development
Intel and AMD have announced a collaboration to integrate their respective x86 CPU optimizations for PyTorch into the open-source deep learning library oneDNN, providing developers with a unified software path and reducing software fragmentation in the x86 AI ecosystem. AMD is integrating its ZenDNN optimizations into oneDNN and joining the UXL Foundation as a member of its steering committee.
Intel and AMD Team Up to Simplify x86 CPU AI Development
Intel and AMD have announced a collaboration to integrate their respective x86 CPU optimizations for PyTorch into the open-source deep learning library oneDNN. This provides developers with a unified software path, reducing software fragmentation within the x86 AI ecosystem.
AMD is integrating its ZenDNN optimizations into oneDNN and joining the UXL Foundation as a member of its steering committee. The two companies are also collaborating on a next-generation set of AI computing extension instructions for x86, which will be further integrated into PyTorch through oneDNN in the future.
Intel China: https://mp.weixin.qq.com/s/JcEFssHmhyL0drohofQNkg
