Cuda Ecosystem

CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs). With CUDA, developers are able to dramatically speed up computing applications by harnessing the power of GPUs.

Community Repos

Experts

Visit Website View on GitHub

Created by: NVIDIA

Released: June 23, 2007

Keywords

gpu 159 python 89 pytorch 69 cpp 58 nvidia 57 machine-learning 50 deep-learning 48 opencl 36 gpgpu 31 rust 31

Languages

C++ 132 Python 122 Cuda 63 Rust 30 Jupyter Notebook 26 C 26 Shell 16 Dockerfile 12 Julia 11 Go 6

Licenses

MIT 158 APACHE-2.0 73 OTHER 55 GPL-3.0 41 BSD-3-CLAUSE 24 GPL-2.0 6 BSD-2-CLAUSE 5 CC0-1.0 4 LGPL-3.0 4 AGPL-3.0 2

https://github.com/Bruce-Lee-LY/decoding_attention

Decoding Attention is specially optimized for multi head attention (MHA) using CUDA core for the decoding stage of LLM inference

14 Aug 2024 14

https://github.com/bkraad47/fat_llama

fat_llama is a Python package for upscaling audio files to FLAC or WAV formats using advanced audio processing techniques

21 Jul 2024 5

Infero

An easy to use, high performant CUDA powered LLM inference library

05 Jun 2024 12

LuisaCompute

High-Performance Rendering Framework on Stream Architectures

20 Nov 2020 636

light-the-torch

Install PyTorch distributions with computation backend auto-detection

09 Jul 2020 218

https://github.com/koide3/fast_gicp

A collection of GICP-based fast point cloud registration algorithms

05 Feb 2020 1,216

computeWorks_examples

Matrix multiplication example performed with OpenMP, OpenACC, BLAS, cuBLABS, and CUDA

16 May 2019 6

nvml_examples

Examples showing how to utilize the NVML library for GPU monitoring

09 May 2019 16

https://github.com/spcl/dace

DaCe - Data Centric Parallel Programming

26 Feb 2019 469

https://github.com/arbor-sim/arbor

The Arbor multi-compartment neural network simulation library

03 Oct 2016 108

TIGRE

TIGRE: Tomographic Iterative GPU-based Reconstruction Toolbox

13 Jun 2016 561