PulseAugur
EN
LIVE 07:34:04
中文(ZH) 摩尔线程完成智谱GLM-5.2 Day-0适配

Moore Threads Adapts Zhipu AI's GLM-5.2 Model on MTT S5000 GPU

Moore Threads has announced the rapid adaptation of Zhipu AI's new flagship open-source model, GLM-5.2, on their MTT S5000 AI training and inference GPU card. This adaptation builds upon previous optimizations for GLM-5.1, focusing on enhancing the MTT S5000's throughput for long-context prefill stages with GLM-5.2's extended context and complex inference demands. The announcement also touches upon AliExpress's new seller program for the US market and unrelated news about SpaceX's market valuation. AI

IMPACT Enhances GPU performance for specific AI models, potentially improving inference speeds for Zhipu AI's ecosystem.

RANK_REASON This is a hardware company adapting an existing AI model, not a new model release from a frontier lab.

Read on 36氪 (36Kr) →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

COVERAGE [1]

  1. 36氪 (36Kr) TIER_1 中文(ZH) ·

    MoorThread Completes GLM-5.2 Day-0 Adaptation

    36氪获悉,6月17日,摩尔线程宣布在AI训推一体全功能GPU智算卡MTT S5000上,完成对智谱新一代开源旗舰模型GLM-5.2的Day-0极速适配。此次适配延续了摩尔线程在GLM-5.1长上下文Prefill与P/D异构分离推理场景中的优化积累,并面向GLM-5.2超长上下文与复杂推理负载,进一步释放MTT S5000在长输入Prefill阶段的高吞吐优势。