AMD moves CPU inference optimizations into Intel-backed oneDNN

Intel and AMD are integrating their CPU optimizations for PyTorch into oneDNN, bringing AMD's downstream ZenDNN work into a shared open-source library, Tom's Hardware reported. AMD also announced that it is joining the UXL Foundation as a steering member. The library detects hardware features at runtime and generates optimized code for operations such as matrix multiplication and convolution. The integration gives developers a single backend for CPU inference optimizations across AMD and Intel systems. Although this collaboration focuses on x86 processors, oneDNN also supports Arm and RISC-V.