HPC for AI: from 10K-GPU clusters and high-speed interconnects to distributed training infrastructure; AI for HPC: accelerating scientific discovery with AI×Science. Reproducible, verifiable AI/HPC engineering practice, continuously open-sourcing scientific software.
Browse the full 15-chapter knowledge system →
An AI infrastructure engineer. Provided InfiniBand high-speed networking and cluster support for a carrier-grade 10K-GPU AI training cluster in Shanghai; led full-stack delivery of large AI/HPC clusters for leading universities and research institutes. Experienced in AI platform & toolchain development, Slurm heterogeneous scheduling, performance benchmarking and system tuning, cloud-native and DevOps automation.
Selected Works: AI-HPC.org · AI-HPC Assistant · AIInfra Knowledge Base · Technical Blog Series
Contact: openaihpc@gmail.com · github.com/openaihpc · Full profile →
AI cracks millennium-level math, a trillion-parameter model goes open-weight, compute moves to space, and "model-defined chips" ship.
Read →GTC Berlin: "AI is manufacturing" — DSX POD factory delivery compresses on-site deployment from months to 14 days.
Read →LingSheng tops TOP500, a 100K-GPU all-domestic cluster goes live, delivered AI datacenter capacity passes 24 GW.
Read →Browse open-source tools, benchmarks and site source; contribute via Issues / PRs.
github.com/openaihpc →Knowledge-base updates, AI×Science features and scientific software releases — technical content only.
Subscribe →Join systems / telemetry / scientific-intelligence groups for online seminars and collaboration.
Learn about working groups →Submit articles, cases or BoF topics — see the contribution guide and benefits.
Contribution guide →