8000 geohot (George Hotz) / Starred · GitHub
[go: up one dir, main page]

Skip to content
View geohot's full-sized avatar

Highlights

  • Pro

Block or report geohot

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Showing results

Wake and Break: a PSP exploit for AMD navi31 (RX 7900) that makes the card accept unsigned IP firmware. Not persistent: a cold power cycle restores stock.

Python 29 Updated Oct 26, 2026

DeepInfer: High-performance, large-scale inference framework for DeepSeek models

Python 26,092 3,207 Updated Nov 3, 2026

PokéBench: local and frontier models play Pokémon Red as a benchmark, with a leaderboard and live stream

HTML 25 2 Updated Nov 15, 2026

The Modular Platform (includes MAX & Mojo)

Mojo 30,143 3,248 Updated Nov 16, 2026

A hierarchical inference engine for frontier MoE models: hot experts on the GPU (CUDA, Vulkan), warm ones in RAM, cold ones streamed from NVMe. Written in C, runs on the hardware you already own.

C 51,449 6,219 Updated Nov 16, 2026

slime is an LLM training framework for RL Scaling.

Python 8,828 1,364 Updated Nov 5, 2026

KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

Jupyter Notebook 1,331 198 Updated Mar 24, 2026

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

233,542 23,224 Updated Apr 20, 2026

A kernel library written in tilelang

Python 2,074 238 Updated Oct 14, 2026
C 133 28 Updated Sep 13, 2026

OpenCode 2 plugin that uses your existing Claude Code credentials — no separate login needed.

TypeScript 1,325 198 Updated Oct 16, 2026

The open source coding agent.

TypeScript 227,134 30,765 Updated Nov 16, 2026

Open-source CUDA, Triton and HIP compiler targeting multiple GPU and CPU architectures.

C 1,757 92 Updated Oct 26, 2026

Documentation for the Mainboard and printable mechanical parts in the Framework Desktop

OpenSCAD 334 23 Updated Jul 6, 2026

A project trying to build a hoverboard controller without semiconductors

20 1 Updated Nov 24, 2025

Code, labs, and resources for O'Reilly AI Systems Performance Engineering: GPU optimization, distributed training, inference scaling, and full-stack tuning.

Python 2,216 314 Updated Nov 16, 2026

A machine learning accelerator core designed for energy-efficient AI at the edge.

Emacs Lisp 2,700 377 Updated Nov 13, 2026

The best ChatGPT that $100 can buy.

Python 60,592 8,506 Updated Sep 7, 2026

Memory Optimizations for Deep Learning (ICML 2023)

Python 121 16 Updated Mar 13, 2024

Exocompilation for productive programming of hardware accelerators

Python 747 61 Updated Nov 12, 2026

Mirage Persistent Kernel: Compiling LLMs into a MegaKernel

Cuda 2,617 277 Updated Nov 15, 2026
Rocq Prover 382 13 Updated Sep 20, 2025

Minimal reproduction of DeepSeek R1-Zero

Python 13,257 1,570 Updated Feb 27, 2026

NanoGPT (124M) in 90 seconds

Python 6,042 928 Updated Sep 28, 2026

Open-source high-performance RISC-V processor

Scala 7,408 984 Updated Nov 16, 2026

Type annotations and runtime checking for shape and dtype of JAX/NumPy/PyTorch/etc. arrays. https://docs.kidger.site/jaxtyping/

Python 1,881 105 Updated Oct 25, 2026

the official Rust and C implementations of the BLAKE3 cryptographic hash function

Assembly 6,529 502 Updated Nov 12, 2026

Minimalistic, extremely fast, and hackable researcher's toolbench for GPT models in 307 lines of code. Reaches <3.8 validation loss on wikitext-103 on a single A100 in <100 seconds. Scales to large…

Python 361 27 Updated Jul 29, 2024
Next
0