CUTLASS: Fast Linear Algebra in CUDA C++ | NVIDIA Technical Blog
Update May 21, 2018: CUTLASS 1.0 is now available as Open Source software at the CUTLASS repository. CUTLASS 1.0 has changed substantially from our preview release described in the blog post below.
CUTLASS: Fast Linear Algebra in CUDA C++ | NVIDIA Technical Blog Technical Blog Subscribe Related Resources AR / VR CUTLASS: Fast Linear Algebra in CUDA C++ Dec 05, 2017 By Andrew Kerr , Duane Merrill , Julien Demouth and John Tran Like Discuss (13) L T F R E AI-Generated Summary Like Dislike CUTLASS is a collection of CUDA C++ templates for implementing high-performance General Matrix Multiplication (GEMM) computations on NVIDIA GPUs. CUTLASS decomposes GEMM into fundamental components, allowing programmers to customize and specialize them within their own CUDA kernels for efficient dense lin
saved by
related reading
- How to Optimize a CUDA Matmul Kernel for cuBLAS-like Performance: a Worklogsiboehm.com
- Reverse-Engineering cuBLASaccu.org
- GitHub - wangzyon/NVIDIA_SGEMM_PRACTICE: Step-by-step optimization of CUDA SGEMMgithub.com
- Outperforming cuBLAS on H100: a Worklogcudaforfun.substack.com
- Inside NVIDIA GPUs: Anatomy of high performance matmul kernels - Aleksa Gordićaleksagordic.com
- Worklog: Optimising GEMM on NVIDIA H100 for cuBLAS-like Performance (WIP) – Hamza's Bloghamzaelshafie.bearblog.dev
- GPU Performance Background User's Guide - NVIDIA Docsdocs.nvidia.com
- Mini Project: GPU Accelerated Matrix Multiplication (almost) like cuBLAS0mean1sigma.com
- Outperforming cuBLAS on H100: a Worklogcudaforfun.substack.com
- Implementing a fast Tensor Core matmul on the Ada Architecture | spatters.caspatters.ca
- Learning CUDA by optimizing matrix-vector multiplication (SGEMV) for cuBLAS-like performance - A worklog – Maharshi's blogmaharshi.bearblog.dev
- Grouped GEMM for Imbalanced Experts on Blackwell: A WIP Workloggauravjain.bearblog.dev