uxlfoundation/oneDNN

NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.

OpenVINO™ is an open source toolkit for optimizing and deploying AI inference

DeepGEMM: clean and efficient BLAS kernel library on GPU

A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downstream deployment frameworks like TensorRT-LLM, TensorRT, vLLM, etc. to optimize inference speed.

CUDA Templates and Python DSLs for High-Performance Linear Algebra

Development repository for the Triton language and compiler

You like pytorch? You like micrograd? You love tinygrad! ❤️

Tile primitives for speedy kernels

FlashMLA: Efficient Multi-head Latent Attention Kernels