Periodic Labs beats frontier models on materials analysis with just 1,300 H200s

LiamFedus · x · 2026-09-16

Periodic Labs (with ex-OpenAI's Liam Fedus) detailed its AI infrastructure: starting from open weights, a mid-training plus RL run peaking at only 1,300 H200 GPUs produced a model called Neon that Pareto-dominates GPT-6 Astra and Claude Fable 5.1 on its X-ray diffraction benchmark. The stack builds on Megatron, SGLang, Miles and Ray—delivering 4.1x Megatron training throughput and 2.5x faster inference—plus custom sandboxing and 95%+ cluster utilization shared with scientific simulation workloads. High-throughput labs in Menlo Park close the loop: fresh experimental data trains models that pick the next experiments, targeting superconductors, magnets and semiconductors.

Related event: Periodic Labs Trains Neon on Just 1,300 H200s to Beat GPT-6 Astra in Materials Science(65 posts)→

Original post →

More from Infra

Infra channel →