ROCm / composable_kernelLinks
Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators
☆401Updated this week
Alternatives and similar repositories for composable_kernel
Users that are interested in composable_kernel are comparing it to the libraries listed below
Sorting:
- collection of benchmarks to measure basic GPU capabilities☆376Updated 3 months ago
- OpenAI Triton backend for Intel® GPUs☆187Updated this week
- Several optimization methods of half-precision general matrix multiplication (HGEMM) using tensor core with WMMA API and MMA PTX instruct…☆415Updated 8 months ago
- An unofficial cuda assembler, for all generations of SASS, hopefully :)☆502Updated 2 years ago
- Assembler for NVIDIA Volta and Turing GPUs☆218Updated 3 years ago
- CUDA Kernel Benchmarking Library☆650Updated last week
- Development repository for the Triton language and compiler☆122Updated this week
- Shared Middle-Layer for Triton Compilation☆251Updated this week
- ☆208Updated 10 months ago
- A Fusion Code Generator for NVIDIA GPUs (commonly known as "nvFuser")☆329Updated this week
- CUDA Matrix Multiplication Optimization☆188Updated 10 months ago
- Yinghan's Code Sample☆329Updated 2 years ago
- A Easy-to-understand TensorOp Matmul Tutorial☆359Updated 8 months ago
- ROCm Communication Collectives Library (RCCL)☆337Updated this week
- AI Tensor Engine for ROCm☆201Updated this week
- A tool for generating information about the matrix multiplication instructions in AMD Radeon™ and AMD Instinct™ accelerators☆95Updated last week
- AMD's graph optimization engine.☆220Updated this week
- Experimental projects related to TensorRT☆105Updated this week
- Stretching GPU performance for GEMMs and tensor contractions.☆242Updated last week
- Step-by-step optimization of CUDA SGEMM☆327Updated 3 years ago
- Development repository for the Triton-Linalg conversion☆189Updated 3 months ago
- A model compilation solution for various hardware☆437Updated 3 weeks ago
- Optimizing SGEMM kernel functions on NVIDIA GPUs to a close-to-cuBLAS performance.☆353Updated 5 months ago
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆365Updated this week
- ☆146Updated this week
- hipBLASLt is a library that provides general matrix-matrix operations with a flexible API and extends functionalities beyond a traditiona…☆97Updated this week
- A simple high performance CUDA GEMM implementation.☆374Updated last year
- ☆109Updated 3 weeks ago
- A tool for bandwidth measurements on NVIDIA GPUs.☆443Updated last month
- BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.☆619Updated 3 weeks ago