浪潮信息发布CPU原生液冷整机柜及超节点，为智能体时代重构算力底座
智能体成规模落地需要专用算力，浪潮推出CPU原生液冷整机柜和低延迟超节点，为多智能体协同重构硬件基础。
随着AI应用从“做大模型”转向“用智能体”，智能体间协同所需的海量CPU算力成为瓶颈。浪潮信息推出业界首款CPU原生液冷整机柜服务器，单柜可支持超4万个智能体同时运行；同时升级元脑SD200超节点，万亿参数模型单Token生成延迟降至4.77ms，为规模化群智协作提供优化基础设施。
正文摘录
.jpg) Editor: Zenan AI agents now have dedicated compute power. Last Thursday, at the Open Computing Technology Conference (OCTS26) held in Beijing, Inspur Information launched a series of AI infrastructure products, including the industry’s first CPU-native liquid-cooled rack server and the YuanBrain SD200 super node AI server.  As the industry consensus shifts from "building large models" to "using agents," Inspur Information has fundamentally restructured the compute infrastructure to address the ev…