Optimized FP16/BF16 x FP4 GPU kernels for AMD GPUs

FlashMLA: Efficient Multi-head Latent Attention Kernels

🚀🚀 Efficient implementations of Native Sparse Attention

The best ChatGPT that $100 can buy.

中国科学技术大学数字电路实验入门指南，2022年由马子睿助教创建。本仓库旨在让各位后续助教能够不断对其进行迭代

CUDA Templates and Python DSLs for High-Performance Linear Algebra

📚LeetCUDA: Modern CUDA Learn Notes with PyTorch for Beginners🐑, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.🎉

🥢像老乡鸡🐔那样做饭。主要部分于2024年完工，非老乡鸡官方仓库。文字来自《老乡鸡菜品溯源报告》，并做归纳、编辑与整理。CookLikeHOC.

在融合ArkFS+vivo50后，加入IPFS/Filecoin的分布式存储设计

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

A high-throughput and memory-efficient inference and serving engine for LLMs

诺亚盘古大模型研发背后的真正的心酸与黑暗的故事。

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

Development repository for the Triton language and compiler

中科大计算机学院部分课程的试卷

a SymDrive copy

2025年12月更新，目前国内可用Docker镜像源汇总，DockerHub国内镜像加速列表，🚀DockerHub镜像加速器

Fast and memory-efficient exact attention

Flash Attention in ~100 lines of CUDA (forward pass only)

Stanford computer networking lab, an elegant TCP/IP implementation

存放中国科大2023春季学期《数字图像处理与分析》课程部分资源

Linux kernel source tree

A wrapper script to build whole-program LLVM bitcode files

[ICLR'25] Fast Inference of MoE Models with CPU-GPU Orchestration

Stanford CS149 -- Assignment 1

中国科学技术大学龙芯杯参赛作品仓库合集

Fully open reproduction of DeepSeek-R1