Autoresearch for GPU kernels. Give it any PyTorch model, go to sleep, wake up to optimized Triton kernels.

Generating Efficient AI-Centric Kernels

A lightweight, general-purpose framework for evaluating GPU kernel and benchmark.

Open source skill library for AI coding agents to write, optimize, and debug high performance compute kernels across CUDA, Triton, and quantized workloads.

Custom Linux kernels purpose-built for Apple Mac hardware

Repo containing artifacts for Neurips 2025 tutorial- How to Build Agents to Generate Kernels for Faster LLMs (and Other Models!)

Fastest MoE/LLM inference runtime for consumer and edge Blackwell GPUs. SN74 on Gittensor.

Agent-queryable ROCm kernel optimization knowledge base for AMD Instinct MI300/gfx942 and MI350/MI355X/gfx950, packaged for Codex CLI and Claude Code with merged-PR provenance, real-silicon validation, and a maintainer-controlled pull-request evidence pipeline.

Evidence-driven CUDA, CUTLASS, Triton and GPU workload optimization for ChatGPT · 使用 ChatGPT 驱动 GPU workload 性能优化

Custom AWS Transform agent that migrates PyTorch/Triton kernels to AWS Trainium NKI (@nki.jit) and compiles, numerically verifies, and profiles every candidate on a real Trainium device before opening a PR.

Extended TileLang as a unified DSL to enable high-performance kernel development for Near-Memory Computing, Distributed Memory AI Accelerators, and Networked Accelerators.

Noeris — autonomous kernel fusion discovery + Triton autotuning for LLM kernels and Gemma layer deeper fusion (A100/H100 wins).

METAL-SCI: a scientific compute benchmark for evolutionary LLM kernel search on Apple Silicon Metal

Compiler MVP that detects Transformer fusion patterns, generates optimized CUDA kernels with WMMA Tensor Cores, and executes them on real GPU hardware — 10.5 TFLOPs on RTX 2070, correctness validated against PyTorch.

可验证的 CUDA 学习主线：SGEMM、通用 GPU 算子、性能优化与轻量推理组件

Reward-hardened evaluation for LLM-generated GPU kernels, built on KernelBench.

RWKV-7 FP8 quantized inference - 6.4x decode speedup on Blackwell GPUs with <0.3% accuracy loss. Full FP8 E4M3 weight quantization with fused Triton kernels.

Automatic Triton kernel generation and optimization for Intel GPU, powered by Claude Code.

Measurement harness for the sliding window attention premium in the vLLM TPU Ragged Paged Attention v3 kernel: per layer decode cost, block size control, throughput, and goodput for Gemma 4 31B on TPU v6e.

Hand-tuned NVIDIA SASS kernels for RTX 3070 Ti (GA104, sm_86): 41,721 dense-equiv 2:4 sparse HGEMM, 11,453 GFLOPS Flash Attention, no cuBLAS / cuDNN / PyTorch. Includes cuasmR, a CRAN-ready R package for cubin read/write + GPU benchmark measurement. 6-chapter tutorial + Chladni-pattern memory layout study.