Efficient Triton Kernels for LLM Training

DeepEP: an efficient expert-parallel communication library

The ASPLOS 2025 / EuroSys 2025 Contest Track

Tensors and Dynamic neural networks in Python with strong GPU acceleration

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

Graph-indexed Pandas DataFrames for analyzing hierarchical performance data

Native PyTorch library for quantization and sparsity

Graph Neural Network Library for PyTorch

Development repository for the Triton language and compiler

Low-Level Graph Neural Network Operators for PyG

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.

HPCToolkit performance tools: measurement and analysis components

A memory profiler for NVIDIA GPUs to explore memory inefficiencies in GPU-accelerated applications.

A compressed alternative to matrix multiplication using state-of-the art compression ROBE-Z

High-order Lagrangian Hydrodynamics Miniapp

Table Maker for Modern C++

Reference implementations of MLPerf™ inference benchmarks

Reference implementations of MLPerf™ training benchmarks

Test overhead of CUPTI PC sampling for CUDA 10

🔥 Hugo website builder, Hugo themes & Hugo CMS. No code, build with widgets! 创建在线课程，学术简历或初创网站。

Revisiting, benchmarking, and refining Heterogeneous Graph Neural Networks.