AMD and Intel join hands to simplify AI development on x86 CPUs

Advanced Micro Devices (AMD) and Intel have announced a collaboration to integrate their x86 CPU optimization efforts for PyTorch into oneDNN, an open-source deep learning library. This initiative, recommended by the x86 Ecosystem Advisory Group, seeks to establish a unified software path for optimised AI workloads across AMD and Intel CPUs, reducing fragmentation within the AI ecosystem.

The companies have aligned on oneDNN as the default open-source library for x86 acceleration within PyTorch.

AMD is integrating its ZenDNN optimisations into oneDNN and has joined the Unified Acceleration Foundation (UXL) as a steering member to further this work through open collaboration.

Developers will benefit from a single point of access for CPU optimisations for PyTorch across both AMD and Intel processors. This simplification aims to make building, maintaining, and deploying AI applications easier across the x86 ecosystem.

AMD and Intel are also working together on AI Compute Extensions (ACE), a next-generation instruction set architecture (ISA) for x86. ACE is designed to accelerate matrix multiplication, a core operation in AI workloads, and will be integrated into PyTorch through future oneDNN updates.

"AMD is advancing an open software ecosystem that gives developers greater choice and makes it easier to build high-performance AI applications. By working with the x86 Ecosystem Advisory Group, we’re bringing the industry together around common technologies that simplify development across x86 processors," said Robert Hormuth, corporate vice president (architecture and strategy), AMD.

“Developers shouldn’t have to think about separate software paths to get the best from an x86 CPU," said Jeff McVeigh, vice president and general manager (data center strategic programs), Intel. "Working with AMD through oneDNN gives the PyTorch community a consistent approach for modern processors and a way to bring future capabilities, including AI Compute Extensions, into the framework.”