A series of high-performance GEMM (General Matrix Multiply) implementations Iteratively optimised for H100 GPUs in Pure CUDA.
☆84Feb 18, 2026Updated 8 months ago
Alternatives and similar repositories for h100_gemm
Users that are interested in h100_gemm are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- ☆51Aug 17, 2026Updated 2 months ago
- We aim to redefine Data Parallel libraries portabiliy, performance, programability and maintainability, by using C++ standard features, i…☆60Updated this week
- native blackwell training backward for gated linear recurrences, and a verifier that catches broken llm-generated gpu kernels☆16Nov 3, 2026Updated last week
- Optimizing diffusion for production-ready speeds☆41Jan 10, 2026Updated 10 months ago
- A survey of modern quantization formats (e.g., MXFP8, NVFP4) and inference optimization tools (e.g., TorchAO, GemLite), illustrated throu…☆12Nov 11, 2025Updated last year
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- ☆20Nov 19, 2024Updated last year
- a fast and customizable CUDA int4 tensor core gemm☆16Aug 2, 2024Updated 2 years ago
- ☆27Aug 28, 2025Updated last year
- Fastest kernels written from scratch☆630Aug 15, 2026Updated 2 months ago
- ☆19Mar 29, 2026Updated 7 months ago
- ☆49Oct 24, 2026Updated 2 weeks ago
- AI agent that helps you learn Tinygrad and updates you on the bounties☆14Aug 5, 2025Updated last year
- ☆27Jun 7, 2026Updated 5 months ago
- coding CUDA everyday!☆78Feb 5, 2026Updated 9 months ago
- Open source password manager - Proton Pass • AdSecurely store, share, and autofill your credentials with Proton Pass, the end-to-end encrypted password manager trusted by millions.
- A PyTorch implementation of the GPT-OSS-20B architecture. All components are coded from scratch: RoPE with YaRN, RMSNorm, SwiGLU with cla…☆238Dec 2, 2025Updated 11 months ago
- parallel_primitiove base on luisacompute☆12Apr 19, 2026Updated 6 months ago
- QuintNet is a research-oriented PyTorch framework designed to explore and implement multi-dimensional parallelism strategies for distribu…☆19Feb 3, 2026Updated 9 months ago
- reproducing the flapping airplanes megakernel, in collab with magni☆135Updated this week
- A Python DSL to write Nvidia PTX for Hopper and Blackwell in JAX and PyTorch☆387Jul 9, 2026Updated 4 months ago
- Tile primitives for speedy kernels☆21Sep 13, 2026Updated 2 months ago
- ☆18Mar 29, 2026Updated 7 months ago
- An MLIR-based compiler that takes GPU kernels and compiles them to real hardware instructions. Interactive web visualizer included.☆149Mar 21, 2026Updated 7 months ago
- A Python DSL for Apple Metal GPU compute☆27Jul 23, 2026Updated 3 months ago
- Deploy open-source AI quickly and easily - Special Bonus Offer • AdRunpod Hub is built for open source. One-click deployment and autoscaling endpoints without provisioning your own infrastructure.
- Repository for the CUDA H100 Course☆80Apr 12, 2026Updated 7 months ago
- A Zero-dependency GEMM library for RTX 50 series (SM120). 100% cuBLAS performance.☆19Aug 31, 2026Updated 2 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆157May 29, 2025Updated last year
- IMMA-based **FP8-as-storage** GEMM experiments for Ampere (sm_86 / RTX 3090 Ti).☆28Jan 30, 2026Updated 9 months ago
- Fast MRI reconstruction on CUDA GPUs☆12Dec 30, 2023Updated 2 years ago
- ☆203Aug 9, 2026Updated 3 months ago
- An agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch…☆163Sep 18, 2026Updated last month
- Automatic differentiation for Triton Kernels☆29Aug 12, 2025Updated last year
- reinvented groq LPU, built to run inference on MicroGPT☆52Aug 24, 2026Updated 2 months ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Minimal TPU implementation with 8x8 systolic array and PyTorch integration☆72Jan 26, 2026Updated 9 months ago
- A practical way of learning Swizzle☆45Feb 3, 2025Updated last year
- Qwen3-0.6B megakernel: 527 tok/s decode on RTX 3090 (3.8x faster than PyTorch)☆145Feb 10, 2026Updated 9 months ago
- [COLM-LLA 2026] The official implementation for paper "AdaExplore: Failure-Driven Adaptation and Diversity-Preserving Search for Efficien…☆24Aug 23, 2026Updated 2 months ago
- LLM serving engines that SkyDiscover builds on the fly for specific AI models and hardware: Qwen3.6-35B-A3B, Qwen3.8-Flash-Next and Flint…☆39Nov 4, 2026Updated last week
- Load and run Llama from safetensors files in C☆16Oct 24, 2024Updated 2 years ago
- Reverse engineering notes. Personal reference only. Everything here is a best-guess reconstruction.☆137Aug 11, 2026Updated 3 months ago
Are these results useful?