Our mission is to accelerate open science to solve the real-world challenges!

How?

Our latest news:

2024 March:We presented our new project to "Automatically Compose High-Performance and Cost-Efficient AI Systems with MLCommons' Collective Mind and MLPerf" at the MLPerf-Bench workshop @ HPCA'24 (collaboration with cKnowledge Ltd and MLCommons).

2024.January we are developing a new generation of our open-source workflow automation technology in collaboration with MLCommons members and cKnowledge Ltd to automate MLPerf inference v4.0 submissions across diverse models, software and hardware from Nvidia, Intel, Qualcomm, AMD and other vendors - don't hestiate to contact us via public Discord server for more details.

2023.December we've completed the artifact evaluation for ACM/IEEE MICRO'23 and prototyped the use of the unified CM automation interface to make it easier to re-run and reproduce experiments from published papers. See our report for more details.


Learn about our mission, open science initiatives and open-source developments, check our Collective Knolwedge technology (donated to MLCommons to benefit everyone), try the new Collective Mind Automation language, read joint Nature'23 article, and participate in our open optimization and reproducibility challenges via CK playground organized in collaboration with MLCommons and the community!

Visit our partner for commercial services to co-design high-performance and cost-effective computer systems for AI and other workloads.