AMD stock
On October 20, 2026, AMD and Intel announced a software collaboration to streamline artificial intelligence development on x86 central processing units by consolidating their PyTorch optimizations into oneDNN. Under this initiative, AMD is integrating its ZenDNN optimizations directly into the open--source oneDNN library—the default acceleration backend for PyTorch on x86 architectures—establishing a unified, vendor--neutral execution path for machine learning developers across both AMD and Intel CPUs. Concurrently, AMD expanded its commitment to multi--architecture computing by joining the Unified Acceleration (UXL) Foundation as a steering member.