The best ChatGPT that $100 can buy.

Reproduction study of Grassmann Flows for sequence modeling (arXiv 2512.19428). Shows 22.6% gap vs claimed 10-15%, includes CUDA kernels with 2x speedup.

Statistics on GPUs

An implementation of NeRF acceleration using RTX cores to compute ray-grid intersections

Masked Depth Modeling for Spatial Perception

A Heterogeneous GPU Platform for Chipyard SoC

The Turkish Sieve Methodology: Deterministic Computation of Twin and Cousin Prime Pairs Using an N/6 Bit Data Structure

Sutskever 30 implementations inspired by https://papercode.vercel.app/

TRELLIS (Microsoft's Image-to-3D generator) running on AMD GPUs with ROCm. Includes Gaussian splatting, mesh extraction, and GLB export. Tested on RX 7800 XT.

RISC-V XV6/Linux SoC, marchID: 0x2b

Sample viewer implementing several rendering methods for 3D gaussians using Vulkan API

A fast framework for writing baseline compiler back-ends in C++

Infrastructure for Machine Learning Guided Optimization (MLGO) in LLVM.

PhotonVision is the free, fast, and easy-to-use computer vision solution for the FIRST Robotics Competition.

Rust library to manipulate CUDA fatbinary format

Official code repository of UTMIST's AI^2 Tournament (2024-2025 version)

Official code repository of UTMIST's AI^2 Tournament

A framework that support executing unmodified CUDA source code on non-NVIDIA devices.

Metal-based implementation of D3D11 and D3D10 for macOS / Wine

A complete neural network built entirely in x86 assembly language that learns to recognize handwritten digits from the MNIST dataset. No frameworks, no high-level languages - just pure assembly - ~…

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RN…

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

An open-source, GPU-accelerated physics simulation engine built upon NVIDIA Warp, specifically targeting roboticists and simulation researchers.

Pure LLM agent powered card game w/ TTS and live dashboard.

Allo Accelerator Design and Programming Framework (PLDI'24)

Intel® Open Image Denoise library

FlashInfer: Kernel Library for LLM Serving