AIC++★ 4,092⑂ 1,221 forksApache-2.0
uxlfoundation/oneDNN
ONE DNN: ACCELERATING AI ON CPU
oneDNN provides high-performance AI kernel implementations for x64, AArch64, RISC‑V, and Intel GPUs. It accelerates PyTorch, TensorFlow, OpenVINO, and ONNX Runtime using AVX10, AMX, and vector extensions.
// KEY FEATURES
- Performs convolution operations with AVX10 and AMX acceleration on x64 CPU.
- Supports int8/bf16 arithmetic for energy‑efficient inference on ARM AArch64.
- Optimizes matrix multiplications for RISC‑V using vector extensions.
- Provides low‑latency computations on Intel GPU via integration with oneAPI.
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram