AMD and Intel unify AI optimizations for x86 CPUs in oneDNN
AMD and Intel are consolidating their PyTorch optimizations into the open-source oneDNN library, giving developers a single software path for AI workloads across x86 CPUs from both vendors.
- AMD and Intel will merge PyTorch optimizations into oneDNN
- Goal is a common software path for AI on all x86 CPUs
- oneDNN remains open source
Read next
Hardware