gemm

Fast inference engine for Transformer models

Tuned OpenCL BLAS

cuDNN Frontend is NVIDIA's modern, open-source entry point to the cuDNN library and a growing collection of high-performance open-source kernels.

BLISlab: A Sandbox for Optimizing GEMM

Several optimization methods of half-precision general matrix multiplication (HGEMM) using tensor core with WMMA API and MMA PTX instruction.

🚀🚀🚀 This repository lists some awesome public CUDA, cuda-python, cuBLAS, cuDNN, CUTLASS, TensorRT, TensorRT-LLM, Triton, TVM, MLIR, PTX and High Performance Computing (HPC) projects.

Optimizing SGEMM kernel functions on NVIDIA GPUs to a close-to-cuBLAS performance.

Multi-Threaded FP32 Matrix Multiplication on x86 CPUs

The HPC toolbox: fused matrix multiplication, convolution, data-parallel strided tensor primitives, OpenMP facilities, SIMD, JIT Assembler, CPU detection, state-of-the-art vectorized BLAS for floats and integers

[DEPRECATED] Moved to ROCm/rocm-libraries repo

AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming

Stepwise optimizations of DGEMM on CPU, reaching performance faster than Intel MKL eventually, even under multithreading.

A Flexible and Energy Efficient Accelerator For Sparse Convolution Neural Network

DBCSR: Distributed Block Compressed Sparse Row matrix library

Single file libraries for C/C++

Hand-written NVFP4 W4A16 CUDA kernels for Volta

FP64 equivalent GEMM by the Ozaki scheme with Int8 Tensor Cores

[DEPRECATED] Moved to ROCm/rocm-libraries repo

PyTorch half precision gemm lib w/ fused optional bias + optional relu/gelu