MSCCL++: A GPU-driven communication stack for scalable AI applications

Microsoft Collective Communication Library

Microsoft Automatic Mixed Precision Library

MSCCL Scheduler

NCCL Profiling Kit

Tutel MoE: Optimized Mixture-of-Experts Library, Support GptOss/DeepSeek/Kimi-K2/Qwen3 using FP8/NVFP4/MXFP4

ROCm Communication Collectives Library (RCCL)

Optimized primitives for collective multi-GPU communication

A validation and profiling tool for AI infrastructure

A Study Path for Game Programmer

LaTeX template for dissertations in Peking University

贵校课程资料民间整理

算法竞赛课件分享

Documentation:

Written in C++ and using SDL, The Powder Toy is a desktop version of the classic 'falling sand' physics sandbox, it simulates air pressure and velocity as well as heat.

FMA: A Dataset For Music Analysis

A script engine for "yu-gi-oh!" and sample gui

VS Code in the browser

A fast multi-producer, multi-consumer lock-free concurrent queue for C++11

Accompanying repository for Let's make a DQN / A3C series.

Assignments for CS294-112.

TLP - Optimize Linux Laptop Battery Life

📚 技术面试必备基础知识、Leetcode、计算机操作系统、计算机网络、系统设计

🌟 Wiki of OI / ICPC for everyone. （某大型游戏线上攻略，内含炫酷算术魔法）

Acm Cheat Sheet

Algorithm Problems I solved... as a hobby

ACMer 入门级算法模板