Alibaba Cloud releases Lingjun Zhenwu M890 super-node instance, capable of handling trillion-parameter-level MoE large model inference on a single machine
At the 2026 World Artificial Intelligence Conference, Alibaba Cloud launched the Lingjun Zhenwu M890 super-node instance, providing AI computing services i...
产品更新IT之家(RSS)
Today AI Intelligence Brief
At the 2026 World Artificial Intelligence Conference, Alibaba Cloud launched the Lingjun Zhenwu M890
super-node instance, providing AI computing services in a super-node form via the public cloud for the first time, and has opened for invitation-based testing in Ulanqab. This instance supports FP8/FP4 low-precision computing, with inter-card connectivity increased to 800GB/s. Training performance is three times higher than the previous generation Zhenwu 810E, and a single unit can handle MoE large model inference at the scale of trillions of parameters. The Zhìsuàn Lingjun single cluster supports up to 130,000 heterogeneous cards, and the instance maintains an average availability of 99.7%.
The readable text on this page was extracted from the public source and organized with attribution, publication time and the original link. Copyright remains with the original author and publisher.
At the 2026 World Artificial Intelligence Conference, Alibaba Cloud launched the Lingjun Zhenwu M890 super-node instance, providing AI computing services i...
IT之家(RSS)2026-07-18T12:47:00.000Z
Scan to open this article
Aioga aggregates global AI updates and preserves source information for verification and citation.