oneDNN
Open-source library of optimized deep learning operations (matmul, convolution, attention) for CPUs (x64, AArch64, RISC-V) and Intel GPUs. Used by PyTorch, TensorFlow, OpenVINO, and ONNX Runtime.
- 4.1k
- C++
- Apache-2.0
- 1 d ago
Open-source library of optimized deep learning operations (matmul, convolution, attention) for CPUs (x64, AArch64, RISC-V) and Intel GPUs. Used by PyTorch, TensorFlow, OpenVINO, and ONNX Runtime.
C++ image processing and machine learning library with using of SIMD: SSE, AVX, AVX-512, AMX for x86/x64, NEON, SVE for ARM, HVX for Hexagon
Library for specialized dense and sparse matrix operations, and deep learning primitives.
A bare-metal C11 inference runtime for Apple Silicon. Bypasses the "Abstraction Tax" by leveraging reverse-engineered AMX instructions and manual register tiling to achieve 34+ tok/s on 1.1B models (7.8x speedup over Python/Accelerate).
Runnable Ambiorix (AMX) / libamxc examples: list, hash table, dynamic string, variant, array, set. Pure C, no C++ runtime.