AI-HPC.orgHPC for AI · AI for HPC
Reproducible engineering practice for AI infrastructure and scientific computing — from 10K-GPU clusters to AI×Science.
Reproducible engineering practice for AI infrastructure and scientific computing — from 10K-GPU clusters to AI×Science.
Hardware and interconnect, storage, Slurm, distributed training, inference engines and AI×Science — engineering notes you can check against a real cluster.
Start reading →AI FOR SCIENCESurrogate models, ML potentials, AI weather and reproducible workflows — the place where AI compute and supercomputing share one floor.
Explore AI×Science →10K-GPU CLUSTERSInfiniBand fabric, Slurm scheduling, benchmarking and tuning — delivered on carrier-grade AI training clusters and university AI/HPC platforms.
See deliveries →