Sampling profiler for Python programs

Cloudflare Turnsile Bypass Lib

🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"

Mirage: Automatically Generating Fast GPU Kernels without Programming in Triton/CUDA

[EuroSys'24] Minuet: Accelerating 3D Sparse Convolutions on GPUs

Allo: A Programming Model for Composable Accelerator Design

A Easy-to-understand TensorOp Matmul Tutorial

TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing. By providing a higher-level interface, algorithm developers can de…

KernelBench: Can LLMs Write GPU Kernels? - Benchmark with Torch -> CUDA problems

Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

ASPLOS'24: Optimal Kernel Orchestration for Tensor Programs with Korch

BladeDISC is an end-to-end DynamIc Shape Compiler project for machine learning workloads.

Development repository for the Triton language and compiler

Python tool for converting files and office documents to Markdown.

ML models + benchmark for tabular data classification and regression

Ahead of Time (AOT) Triton Math Library

Cataloging released Triton kernels.

Exocompilation for productive programming of hardware accelerators

Simplify Caddy configs with SSL, proxies, file servers, security headers, compression & more.

Shared Middle-Layer for Triton Compilation

triton ops

EquiTriton is a project that seeks to implement high-performance kernels for commonly used building blocks in equivariant neural networks, enabling compute efficient training and inference.

Puzzles for learning Triton

Fast low-bit matmul kernels in Triton

Ongoing research training transformer models at scale