Topics / #gpu-programming

GPU programming projects on GitHub

Programming GPUs for parallel computing, simulation, and performance profiling.

As of , the most-starred GPU programming projects on GitHub are taichi (28.4K stars), codon (16.9K stars) and gpu.js (15.5K stars). GitHub.Locker tracks 28 GPU programming repositories with 214.1K stars combined. The fastest-growing on October 16 was KytyPS5, with +156 stars in a day.

28Projects
214.1KStars in total
+274Stars in a day
Browse in the catalog →github.com/topics/gpu-programming ↗

Hot right now

Most stars gained in a day · Oct 16

+156

KytyPS5

KytyPS5/KytyPS5 · ★ 8.7K

KytyPS5 is a free and open-source PlayStation 5 emulator (translation layer) written in C++ for Windows and Linux, with experimental macOS support, based on a heavily modified version of the Kyty project.

+29

spirula-studio

harry7557558/spirula-studio · ★ 1.7K

Spirula Studio is a 3D Gaussian Splatting trainer: it turns raw photos or video into splats and textured meshes, all from one self-contained binary.

+15

ZLUDA

vosen/ZLUDA · ★ 15.1K

ZLUDA is a drop-in replacement for CUDA that lets graphics cards from vendors other than NVIDIA run CUDA software.

+15

hells-gate-recomp

florinp93/hells-gate-recomp · ★ 540

It is a static recompilation port that brings the Xbox 360 version of Dante's Inferno to native PC, using the ReXGlue SDK to translate the original PowerPC XEX executable directly into C++ code that runs natively on Windows (D3D12) and Linux (Vulkan).

The all-time top

28 projects · by stars

  1. #1 · taichi-dev/taichi

    taichi

    Taichi Lang is an open-source language for high-performance GPU programming embedded in Python: you write Python-like code and its JIT compiler turns it into efficient parallel GPU or CPU machine code.

    28.4K
  2. #2 · exaloop/codon

    codon

    Codon is a high-performance Python implementation that compiles your code straight to native machine code — typically 10-100x faster than vanilla Python on a single thread, on par with C/C++.

    16.9K
  3. #3 · gpujs/gpu.js

    gpu.js

    GPU.js is a JavaScript library that takes the functions you write and runs them on your GPU as shaders instead of on the CPU — a big speedup for parallel math like matrix multiplication.

    15.5K
  1. 4

    ZLUDA

    15.1Kvosen/ZLUDA · Rust

    ZLUDA is a drop-in replacement for CUDA that lets graphics cards from vendors other than NVIDIA run CUDA software.

  2. 5

    scalene

    13.5Kplasma-umass/scalene · Python

    Scalene is a high-performance CPU, GPU, and memory profiler for Python. It is also the first profiler to add AI-powered optimization proposals, via providers like OpenAI, Azure, Bedrock, or local Ollama models.

  3. 6

    GPU-Puzzles

    12.5Ksrush/GPU-Puzzles · Jupyter Notebook

    GPU-Puzzles is an interactive Jupyter notebook that teaches beginner GPU programming by having you implement CUDA kernels directly, puzzle by puzzle, instead of reading theory.

  4. 7

    cupy

    12.4Kcupy/cupy · Python

    CuPy is a NumPy/SciPy-compatible array library for GPU-accelerated computing with Python: you write ordinary Python array code, and it executes on NVIDIA or AMD GPUs.

  5. 8

    taskflow

    12.2Ktaskflow/taskflow · C++

    Taskflow is a C++ library that helps you write high-performance task-parallel programs with high programming productivity, letting you build and run graphs of dependent tasks across CPU threads and GPUs.

  6. 9

    LeetCUDA

    12.1Kxlite-dev/LeetCUDA · Cuda

    LeetCUDA is an open-source book of CUDA learning notes for beginners, paired with a library of 200+ hand-written CUDA kernels (elementwise ops, HGEMM, Flash-Attention, etc.) and 100+ CUDA/LLM blog posts.

  7. 10

    sokol

    10.4Kfloooh/sokol · C

    Sokol is a collection of minimal, cross-platform, standalone C header libraries for building apps and games: 3D graphics, windowing and input, audio, time, async file loading and argument parsing.

  8. 11

    DeepEP

    10.3Kdeepseek-ai/DeepEP · Cuda

    DeepEP is a high-performance communication library for MoE (mixture-of-experts) models, providing fast all-to-all GPU kernels for expert-parallel dispatch and combine operations used in ML training and inference.

  9. 12

    Game-Programmer-Study-Notes

    10.1KQianMo/Game-Programmer-Study-Notes

    This is a personal collection of study notes from a game programmer's career, covering computer graphics, real-time rendering, GPU programming, design patterns, and software engineering. It works like an enhanced blog, condensing and summarizing classic industry books and technical topics.

  10. 13

    DeepGEMM

    9Kdeepseek-ai/DeepGEMM · Cuda

    DeepGEMM is a unified CUDA kernel library that provides high-performance tensor core GEMM operations (FP8, FP4, BF16), fused MoE kernels, and attention-indexer kernels used in large language models such as DeepSeek.

  11. 14

    KytyPS5

    8.7KKytyPS5/KytyPS5 · C++

    KytyPS5 is a free and open-source PlayStation 5 emulator (translation layer) written in C++ for Windows and Linux, with experimental macOS support, based on a heavily modified version of the Kyty project.

  12. 15

    tilelang

    8.6Ktile-ai/tilelang · Python

    TileLang (tile-lang) is a domain-specific language for writing high-performance GPU/CPU/NPU kernels like GEMM, Dequant GEMM, FlashAttention, and LinearAttention, using Pythonic syntax backed by a TVM-based compiler.

  13. 16

    lectures

    6.9Kgpu-mode/lectures · Jupyter Notebook

    This is the supplementary material repository for the GPU MODE lectures YouTube channel — a growing collection of notebooks, slides and code from each numbered lecture on GPU and CUDA programming.

  14. 17

    gpgpu-sim_distribution

    1.8Kgpgpu-sim/gpgpu-sim_distribution · C++

    GPGPU-Sim is a cycle-level simulator that models contemporary NVIDIA GPUs running CUDA or OpenCL workloads, letting researchers study GPU microarchitecture without real hardware. It also bundles a visualization tool (AerialVision) and a power model (AccelWattch).

  15. 18

    spirula-studio

    1.7Kharry7557558/spirula-studio · C++

    Spirula Studio is a 3D Gaussian Splatting trainer: it turns raw photos or video into splats and textured meshes, all from one self-contained binary.

  16. 19

    NoGraphicsAPI

    1.6Ksebbbi/NoGraphicsAPI · C++

    NoGraphicsAPI is an experimental Vulkan 1.4 implementation of Sebastian Aaltonen's 'No Graphics API' idea, exploring what disappears from a conventional graphics API when shaders use raw 64-bit GPU pointers instead of buffer objects and bindings.

  17. 20

    TornadoVM

    1.5Kbeehive-lab/TornadoVM · Java

    An open-source GPU programming framework for Java: it JIT-compiles Java bytecode at runtime into NVIDIA CUDA, OpenCL C or Apple Metal code, so Java programs run on GPUs and multi-core CPUs. Works with JDK 21-27.

  18. 21

    tidewater

    1.3Kdgreenheck/tidewater · JavaScript

    Tidewater is an island fishing game that runs right in your browser: cast from the pier, the beach or your own boat, fight the fish, sell the catch to Joe, and spend it on better gear at Marta's chandlery.

  19. 22

    crosstl

    1.1KCrossGL/crosstl · Python

    CrossTL is a shader and compute program translator built around CrossGL, an intermediate representation (IR) language that bridges graphics APIs, GPU-compute platforms, and systems-language targets.

  20. 23

    OpenDLSS-NR

    921maanHimself/OpenDLSS-NR · JavaScript

    OpenDLSS-NR is a from-scratch Vulkan reimplementation of NVIDIA's DLSS 5 Neural Rendering network, matching the original bit-exactly down to every internal block boundary, not just the final image.

  21. 24

    hells-gate-recomp

    540florinp93/hells-gate-recomp · C++

    It is a static recompilation port that brings the Xbox 360 version of Dante's Inferno to native PC, using the ReXGlue SDK to translate the original PowerPC XEX executable directly into C++ code that runs natively on Windows (D3D12) and Linux (Vulkan).

  22. 25

    mcopt

    471noahdunnagan/mcopt · Java

    mcopt is a Minecraft mod that adds a native Metal renderer for Minecraft: Java Edition, aimed specifically at making the game run faster on Apple Silicon Macs.

  23. 26

    d4r

    288countervolts/d4r · C++

    d4r runs NVIDIA's official DLSS Super Resolution library inside Windows games on AMD Radeon GPUs under Linux and Proton, translating the DLSS network through ZLUDA (CUDA on ROCm/HIP) and replacing the heaviest kernels with hand-written RDNA3/RDNA4 code. It supports the DLSS 3 CNN model and the DLSS 4/4.5 transformer models; DLSS 5 is not supported.

  24. 27

    v100-skinny

    172dnv2003/v100-skinny · Python

    v100-skinny provides hand-written NVFP4 (W4A16) and FP8 (W8A16) CUDA kernels that let four Tesla V100-SXM2-16GB GPUs run Qwen3.8-27B's published mixed NVFP4/FP8 weights directly, serving the model through an OpenAI-compatible API.

  25. 28

    ComfyUI-DLSS5-NR

    172lisitskyaa/ComfyUI-DLSS5-NR · C++

    It is an unofficial, experimental ComfyUI custom node that integrates NVIDIA DLSS 5 Neural Rendering (NGX feature 18) directly into image-generation workflows.

Often tagged together