NVIDIA unveils modular DSX POD AI factory design
NVIDIA introduced the DSX POD reference design at GTC Berlin, packaging seven Vera Rubin racks, cabling, cooling connections, and other infrastructure into a modular container that is integrated and validated before delivery. NVIDIA says this can reduce on-site deployment at a ready facility from six months to as little as 14 days. The company is working with Dell Technologies, HPE, Lenovo, Schneider Electric, and Vertiv on global deliveries planned for early 2027.
The reference design turns AI data-center deployment into a largely pre-integrated factory process, potentially shortening the path from equipment delivery to operational Vera Rubin capacity.
英伟达推出模块化AI工厂设计方案,部署时间可从6个月缩短至14天-36氪
登录
搜索
#NVIDIA#DSX POD#Vera Rubin#Dell Technologies#HPE#Lenovo#Schneider Electric#Vertiv