Open-source gpu-computing software

386 open-source libraries, tools and applications tagged gpu-computing, ranked by how closely gpu-computing is their primary purpose, then by adoption and maintenance health. Each entry links to a page with sources, use cases and when to choose or avoid it.

386 products, primary matches first, then adoption-weighted; health v2 shown.

Product	Health v2	Stars	Maturity
learning-at-home/hivemind Hivemind is a PyTorch library for decentralized deep learning across the Internet, enabling training of large models on hundreds of volunte…	53	2541	active
mflux-community/mflux MFLUX is a native MLX implementation of state-of-the-art generative image and video models (Flux, Qwen-Image, Z-Image, and others), ported …	91	2526	active
Alibaba-Quark/LiveAvatar LiveAvatar is an open-source implementation of an ECCV 2026 paper for streaming, real-time, infinite-length audio-driven avatar video gener…	58	2456	active
NovaSky-AI/SkyRL SkyRL is a modular full-stack reinforcement learning library for post-training large language models, combining a training framework (skyrl…	83	2436	active
Olive Olive is Microsoft's AI model optimization toolkit for the ONNX Runtime, automating finetuning, conversion, quantization, and compression o…	89	2402	active
apple/axlearn AXLearn is a Python deep learning library built on JAX and XLA for developing and training large-scale models, with an object-oriented conf…	69	2375	active
Selkies Remote Desktop Platform Selkies is an open-source low-latency, GPU/CPU-accelerated Linux remote desktop and game streaming platform that delivers an HTML5 web clie…	98	2309	active
dstack dstack is an open-source, vendor-agnostic control plane for GPU provisioning and orchestration that works across GPU clouds, Kubernetes, an…	95	2277	active
google-deepmind/mujoco_playground MuJoCo Playground is an open-source Python library of GPU-accelerated robot learning environments built on MuJoCo MJX and MuJoCo Warp. It s…	74	2250	active
ByteDance-Seed/VeOmni VeOmni is a PyTorch-native framework for single- and multi-modal model pre-training and post-training, with a modular, trainer-free design …	87	2246	active
rapidsai/cugraph cuGraph is NVIDIA's RAPIDS collection of GPU-accelerated graph analytics libraries, offering Python, C, and C++ APIs for building graphs an…	90	2237	active
hls4ml hls4ml is a Python package that converts machine learning models from Keras, PyTorch, and ONNX into high-level synthesis (C++) code for FPG…	80	2185	active
PrimeIntellect-ai/prime-rl prime-rl is a Python framework for large-scale, fully asynchronous reinforcement learning training of language models, built on FSDP2 for t…	88	2164	active
HRM-Text HRM-Text is a 1B-parameter text generation model based on the hierarchical recurrent HRM architecture, released with a complete pretraining…	57	2093	active
vitoplantamura/OnnxStream A lightweight C++ inference library for ONNX models that streams weights to run large models in very little memory, accelerated by XNNPACK.…	56	2089	active
NVIDIA-NeMo/RL NeMo RL is NVIDIA's open-source post-training library for scaling reinforcement learning methods (GRPO, PPO, DPO, SFT, distillation) on LLM…	79	2077	active
marcoslucianops/DeepStream-Yolo A collection of configuration files, parsers, and conversion utilities for running YOLO-family object detection models on NVIDIA DeepStream…	57	2074	active
DeePMD-kit DeePMD-kit is a deep learning package for building many-body potential energy representations and running molecular dynamics simulations. I…	92	2055	stable
OpenMM OpenMM is a high-performance toolkit and library for molecular dynamics simulation, with optimized GPU-accelerated kernels. It can be used …	93	2000	stable
llvm/torch-mlir Torch-MLIR is a compiler project providing first-class translation of PyTorch programs into the MLIR compiler ecosystem. It lets hardware v…	99	1935	active
laugh12321/TensorRT-YOLO A C++/Python deployment toolkit for running YOLO-family models (YOLOv3 through YOLO26) on NVIDIA GPUs using TensorRT, with custom plugins, …	71	1893	active
NVIDIA-AI-IOT/Lidar_AI_Solution NVIDIA's collection of GPU-accelerated Lidar AI inference solutions for autonomous driving, including optimized implementations of PointPil…	69	1885	active
beam-cloud/beta9 Beam (beta9) is an open-source serverless runtime for AI workloads, providing GPU inference endpoints, isolated sandboxes for running untru…	91	1827	active
nvidia-isaac/cuVSLAM cuVSLAM is NVIDIA's CUDA-accelerated library for real-time visual odometry and simultaneous localization and mapping (SLAM). It supports mu…	78	1824	active
kingoflolz/mesh-transformer-jax A JAX/Haiku library implementing model-parallel training and inference of transformer models using xmap/pjit operators, similar to Megatron…	32	6384	maintenance
googlecolab/google-colab-cli A Python-based command-line interface for Google Colab that lets users provision CPU, GPU, and TPU runtimes, execute local scripts and note…	60	1656	active
GoMLX GoMLX is an accelerated machine learning and math framework for Go, comparable to PyTorch/JAX/TensorFlow. It offers differentiable operator…	96	1651	active
Higgsfield Higgsfield is an open-source GPU orchestration and machine learning framework for fault-tolerant, distributed training of very large models…	67	6047	maintenance
UbiquitousLearning/mllm MLLM is a fast, lightweight multimodal LLM inference engine written in C++ for mobile and edge devices, with backends for ARM CPU, Qualcomm…	71	1617	active
xLLM-AI/xllm xLLM is a high-performance C++ inference engine for LLM, VLM, DiT and recommendation models, optimized for heterogeneous AI accelerators su…	80	1600	active
alibaba/Pai-Megatron-Patch Pai-Megatron-Patch is an open-source deep learning training toolkit from Alibaba Cloud for large-scale training and inference of LLMs and V…	50	1599	active
pytorch/FBGEMM FBGEMM is a collection of highly optimized low-precision matrix multiplication and convolution kernels for server-side deep learning infere…	91	1595	active
Lightning-AI/lightning-thunder Lightning Thunder is a source-to-source deep learning compiler for PyTorch that optimizes models for training and inference. It provides a …	70	1475	active
mit-han-lab/torchsparse TorchSparse is a high-performance PyTorch library for sparse convolution on 3D point clouds, with optimized GPU kernels for both training a…	23	1473	active
FeiYull/TensorRT-Alpha A C++/CUDA library providing TensorRT-accelerated deployment for 30+ popular computer vision models including YOLOv3-v8, YOLOv8-Pose/Seg/Cl…	100	1464	active
deepseek-ai/EPLB EPLB is DeepSeek's open-source Expert Parallelism Load Balancer for Mixture-of-Experts models. It computes balanced expert replication and …	23	1443	active
heterodb/pg-strom PG-Strom is a PostgreSQL extension that accelerates SQL analytics and batch workloads using GPU devices, NVMe-SSD storage, and Apache Arrow…	77	1409	active
SpikingBrain SpikingBrain-7B is a brain-inspired large language model that combines hybrid efficient attention, MoE modules, and spike encoding, with a …	51	1408	active
mratsim/Arraymancer Arraymancer is a fast, ergonomic N-dimensional tensor (ndarray) library written in Nim, inspired by NumPy and PyTorch. It provides CPU, CUD…	57	1407	active
erwincoumans/tiny-differentiable-simulator Tiny Differentiable Simulator (TDS) is a header-only C++ and CUDA physics library for rigid-body dynamics with zero dependencies, supportin…	67	1376	active
KhronosGroup/SPIRV-Tools SPIR-V Tools is a Khronos Group project providing an API and command-line tools for processing SPIR-V modules, including an assembler, bina…	95	1371	stable
uxlfoundation/scikit-learn-intelex Intel's Extension for Scikit-learn is a free AI accelerator that speeds up existing scikit-learn workflows on CPUs and GPUs, claiming up to…	87	1356	active
hao-ai-lab/LookaheadDecoding A Python library implementing Lookahead Decoding, an exact parallel decoding algorithm that accelerates LLM inference without a draft model…	28	1343	active
mapillary/inplace_abn A PyTorch extension library implementing In-Place Activated BatchNorm (InPlace-ABN), which redefines BN plus nonlinear activation as a sing…	67	1333	stable
ai-dynamo/nixl NVIDIA Inference Xfer Library (NIXL) is a C++/Python library that accelerates point-to-point communication in AI inference frameworks like …	87	1312	active
plaidml/plaidml PlaidML is a portable tensor compiler that enables deep learning on hardware (especially GPUs and embedded devices) not well supported by m…	10	4566	maintenance
ModelCloud/GPTQModel GPTQModel is a production-ready Python toolkit for quantizing (compressing) large language models using GPTQ, AWQ, and related methods, wit…	92	1274	active
dendenxu/fast-gaussian-rasterization A drop-in replacement for diff-gaussian-rasterization that renders 3D Gaussian Splatting scenes using a geometry-shader-based GPU pipeline …	16	1215	active
Zefan-Cai/R-KV R-KV is a training-free, redundancy-aware KV cache compression method for reasoning LLMs, discarding repetitive tokens on-the-fly during de…	64	1214	active
Cerebras/modelzoo Cerebras Model Zoo is a collection of reference deep learning model implementations (Llama, Mixtral, DINOv2, Llava, etc.) with configs and …	74	1202	active
steelbrain/ffmpeg-over-ip A client/server tool that lets applications use GPU-accelerated ffmpeg on a remote machine over a single TCP connection, without GPU passth…	78	1187	active
MoonshotAI/MoonEP MoonEP is an Expert Parallelism communication library for Mixture-of-Experts training that keeps token loads perfectly balanced across rank…	56	1146	active
flagos-ai/FlagGems FlagGems is a high-performance operator library for large language models written in the Triton language, providing backend-neutral GPU ker…	88	1137	active
Omniverse Kit Sample Suite A template toolkit from NVIDIA for building GPU-accelerated, OpenUSD-based 3D applications with the Omniverse Kit SDK. It provides pre-conf…	74	1106	active
aiptimizer/TurboOCR TurboOCR is an extremely fast GPU-accelerated document parser written in C++ that combines OCR, layout analysis, table extraction, and form…	80	1104	active
sirius-db/sirius Sirius is a GPU-native SQL analytics engine written in C++ that accelerates query execution by offloading it to GPUs. It integrates with ex…	78	1093	active
NVlabs/Fast-dLLM NVIDIA's official implementation of Fast-dLLM, a family of training-free and fine-tuning-based acceleration techniques for diffusion-based …	54	1090	active
meta-pytorch/monarch Monarch is a distributed programming framework for PyTorch built on scalable actor messaging, with actors grouped into meshes, supervision-…	83	1081	active
HannesStark/boltzgen BoltzGen is an open-source all-atom generative diffusion model for designing protein and peptide binders against arbitrary biomolecular tar…	75	1081	active
FINN FINN is an open-source dataflow compiler from AMD/Xilinx that generates highly efficient FPGA accelerators for quantized neural network (QN…	85	1077	active
open-gigaai/giga-train GigaTrain is an efficient and scalable Python training framework for large AI models, supporting distributed multi-GPU/multi-node execution…	59	1072	active
scrya-com/rotorquant RotorQuant is a KV cache compression method for LLM inference that replaces full d×d orthogonal rotations with block-diagonal Clifford roto…	47	1049	active
NVIDIA-NeMo/Skills Nemo-Skills is a collection of Python pipelines for improving the skills of large language models, covering synthetic data generation, mode…	69	1042	active
microsoft/Tutel Tutel is Microsoft's optimized Mixture-of-Experts (MoE) library for efficient training and inference of large language models, featuring dy…	79	1028	active
fla-org/native-sparse-attention Efficient Triton kernel implementations of Native Sparse Attention (NSA), a hardware-aligned, natively trainable sparse attention mechanism…	47	1023	active
facebookresearch/fairscale FairScale is a PyTorch extension library providing composable modules and APIs for high-performance, large-scale distributed training, incl…	27	3403	maintenance
XYZ-AI-Lab/axrl AxisRL is an agentic reinforcement learning post-training framework for large language models, built on SGLang for high-throughput rollout …	53	984	active
isaac-sim/IsaacGymEnvs A collection of example reinforcement learning environments for NVIDIA Isaac Gym, a GPU-accelerated physics simulator. It provides a Gym-st…	10	2952	maintenance
nerdyrodent/VQGAN-CLIP A Python application for running VQGAN+CLIP text-to-image generation locally on your own GPU, derived from Katherine Crowson's Google Colab…	32	2645	maintenance
IST-DASLab/gptq Reference implementation of GPTQ, a one-shot post-training weight quantization method for large generative transformer models based on appr…	32	2386	maintenance
mit-han-lab/once-for-all Once-for-All (OFA) is a PyTorch library implementing the ICLR 2020 Once-for-All network, which trains a single supernet that can be special…	23	1953	maintenance
Maratyszcza/NNPACK NNPACK is a C99 acceleration package providing high-performance SIMD and multi-core CPU implementations of neural network layers, especiall…	32	1710	maintenance
bes-dev/stable_diffusion.openvino A Python CLI implementation of Stable Diffusion text-to-image generation optimized for Intel CPUs and GPUs via OpenVINO. It supports text-t…	32	1534	maintenance
BigScience BigScience is a research project training large transformer language models (BERT, GPT-style) at scale, built on a fork of Megatron-LM inte…	32	1448	maintenance
chengzeyi/stable-fast Stable Fast is an ultra lightweight performance optimization framework for HuggingFace Diffusers pipelines on NVIDIA GPUs. It achieves SOTA…	21	1305	maintenance
NVIDIA-AI-IOT/trt_pose trt_pose is a Python library from NVIDIA for real-time human pose estimation accelerated with TensorRT, targeting NVIDIA Jetson and other N…	23	1067	maintenance
microsoft/nnfusion NNFusion is a flexible and efficient deep neural network (DNN) compiler that generates high-performance executables from model descriptions…	10	1005	maintenance
Rust-GPU/rust-gpu A Rust compiler backend that emits SPIR-V, making Rust a first-class language for writing GPU shaders targeting Vulkan. It lets developers …	80	3376	experimental
jephersonRD/Maquina-V5 A collection of Google Colab notebooks and scripts that spin up a temporary cloud gaming PC with an NVIDIA Tesla T4 GPU, install Steam, and…	47	2336	experimental
volcengine/veScale veScale is a PyTorch distributed training library from ByteDance for hyperscale training of large language models and reinforcement learnin…	54	1041	experimental
horovod/horovod Horovod is a distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet, originally developed at Uber. I…	10	14680	abandoned
intel/ipex-llm IPEX-LLM is a PyTorch LLM acceleration library for Intel hardware (iGPU, NPU, Arc/Flex/Max GPUs, and CPU), offering low-bit quantization (F…	10	8845	abandoned
NVIDIA/DIGITS DIGITS is a web application for training deep learning models on GPUs, supporting frameworks like Caffe, Torch, and TensorFlow with a brows…	10	4176	abandoned
Golem (Clay) Clay Golem is a Python implementation of a decentralized peer-to-peer marketplace for renting idle CPU and GPU computing power, with Ethere…	10	2877	abandoned
intel/intel-extension-for-pytorch A Python package extending official PyTorch with Intel-specific optimizations for CPUs and GPUs, including quantization and LLM inference a…	10	2012	abandoned
nod-ai/AMD-SHARK-Studio AMD-SHARK Studio is a web UI distribution for high-performance machine learning inference built on SHARK and IREE, primarily for running St…	45	1451	abandoned