Skip to content

A validation and profiling tool for AI infrastructure

License

Notifications You must be signed in to change notification settings

RyoYang/superbenchmark

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SuperBench

Build Image Codecov Website Latest Release Docker Pulls License

Azure Pipelines Build Status
cpu-unit-test Build Status
cuda-unit-test Build Status
ansible-integration-test Build Status

SuperBench is a validation and profiling tool for AI infrastructure.

📢 v0.10.0 has been released!

Check aka.ms/superbench for more details.

Citations

To cite SuperBench in your publications:

@inproceedings {superbench,
	author = {Yifan Xiong and Yuting Jiang and Ziyue Yang and Lei Qu and Guoshuai Zhao and Shuguang Liu and Dong Zhong and Boris Pinzur and Jie Zhang and Yang Wang and Jithin Jose and Hossein Pourreza and Jeff Baxter and Kushal Datta and Prabhat Ram and Luke Melton and Joe Chau and Peng Cheng and Yongqiang Xiong and Lidong Zhou},
	title = {{SuperBench}: Improving Cloud {AI} Infrastructure Reliability with Proactive Validation},
	booktitle = {2024 USENIX Annual Technical Conference (USENIX ATC 24)},
	year = {2024},
	isbn = {978-1-939133-41-0},
	address = {Santa Clara, CA},
	pages = {835--850},
	url = {https://www.usenix.org/conference/atc24/presentation/xiong},
	publisher = {USENIX Association},
	month = jul
}

Trademarks

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.

About

A validation and profiling tool for AI infrastructure

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 65.3%
  • C++ 22.1%
  • Cuda 5.4%
  • HLSL 2.6%
  • Dockerfile 2.6%
  • CMake 1.0%
  • Other 1.0%