Moore Threads Pretrains 236B MoE on 10,000-GPU Cluster

Moore Threads announced at the WAIC that it pre-trained a 236-billion-parameter MoE model from scratch on a 10,000-GPU domestic cluster, highlighting rapid advancements in China's AI hardware and compute capabilities.

2026-07-22 ~ 2026-07-22 · 2 related posts

• Moore Threads says it pre-trained a 236B MoE model on a 10,000-GPU cluster — pstAsiatech · 2026-07-22
• China’s AI hardware scene now has a 10K-GPU cluster and a 236B MoE model — teortaxesTex · 2026-07-22