Search

LLM Inference Engine for Metal, CUDA and Vulkan.

AI 基础知识 - GPU 架构、CUDA 编程、大模型基础及AI Agent 相关知识。

Agent Skills for NVIDIA products — install into Claude Code, Codex, and other coding agents to run Physical AI, robotics, simulation, CUDA, and RAG workflows end to end.

NVIDIA's CUDA platform for Rust. Host runtime crates plus Tile (cutile-rs) and SIMT (cuda-oxide) kernel programming models in idiomatic Rust.

CUDA Templates and Python DSLs for High-Performance Linear Algebra

Samples for CUDA Developers which demonstrates features in CUDA Toolkit

Don't memorize ML. Understand it from first principles including Inference, Cuda, Deep Learning, Data Science and Reinforcement Learning

CUDA on non-NVIDIA GPUs

CUDA Python: Performance meets Productivity

High-performance LLM, VLM, embedding, reranking, and audio inference for Apple Silicon, NVIDIA CUDA and AMD ROCm systems.

Pure-Rust inference engine for Qwen GGUF models (Qwen3 to Qwen3.8, dense and MoE up to 770B+). CUDA and Vulkan (NVIDIA, AMD), hybrid CPU+GPU experts, MTP speculative decoding, OpenAI and Anthropic compatible server. Up to ~1.7x llama.cpp decode on a 12 GB RTX 3060.

LWJGL is a Java library that enables cross-platform access to popular native APIs useful in the development of graphics (OpenGL, Vulkan, bgfx), audio (OpenAL, Opus), parallel computing (OpenCL, CUDA) and XR (OpenVR, LibOVR, OpenXR) applications.

PopSift is an implementation of the SIFT algorithm in CUDA.

Backend.AI is a streamlined, container-based computing cluster platform that hosts popular computing/ML frameworks and diverse programming languages, with pluggable heterogeneous accelerator support including CUDA GPU, ROCm GPU, Gaudi NPU, Google TPU, GraphCore IPU and other NPUs.

Thin, unified, C++-flavored wrappers for the CUDA APIs

Multi-GPU CUDA stress test

how to optimize some algorithm in cuda.

Vulkan, OptiX and CUDA Interoperation Modular Rendering Library and Framework for PC/Linux/Android

Omega-AI is a Java-based deep learning framework that helps you quickly build neural networks for inference and training. Its engine supports automatic differentiation, multithreading, and GPU acceleration with CUDA and cuDNN.

HIPIFY: Convert CUDA to Portable C++ Code

Antares: an automatic engine for multi-platform kernel generation and optimization. Supporting CPU, CUDA, ROCm, DirectX12, GraphCore, SYCL for CPU/GPU, OpenCL for AMD/NVIDIA, Android CPU/GPU backends.

Up to 100x faster strings for C, C++, CUDA, Python, Rust, Swift, JS, & Go, leveraging NEON, AVX2, AVX-512, SVE, GPGPU, & SWAR to accelerate search, hashing, sorting, edit distances, sketches, and memory ops 🦖

CUDA Core Compute Libraries

CUDA Kernel Benchmarking Library

LYGIA, it's a granular and multi-language (GLSL, HLSL, Metal, WGSL, WEGL and CUDA) shader library designed for performance and flexibility

Lightning fast C++/CUDA neural network framework

chipStar is a tool for compiling and running HIP/CUDA on SPIR-V via OpenCL or Level Zero APIs.

AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.

An Open-source book with Modern CUDA learning Notes for Beginners - BF16/FP8/FP4, HGEMM, CuTe, Flash-Attention, etc.

Ruby gem for running state-of-the-art language models locally. Access LLMs, embeddings, rerankers, and NER models directly from Ruby using Rust-powered Candle with Metal/CUDA acceleration.