Nvidia has introduced DSX POD, a reference design for Vera Rubin AI systems assembled inside modular containers. The company says partners will build, integrate and test the containers before delivery, then connect them to a prepared site.
Nvidia says this approach can cut deployment time at a prepared site from six months to as little as 14 days. That estimate applies to installation after the site is ready, with power, cooling, networking and other supporting infrastructure in place.
Under the design, partners complete rack-to-rack cabling, wiring, piping and functional checks at a manufacturing facility. Nvidia says that work can take place while the destination site is being prepared, shortening the time between delivery and operation.
The containers bring together Vera Rubin computing systems and the infrastructure needed to run them. Nvidia says a faster installation could let operators begin model training and production inference sooner. It also says operators could start generating production tokens and earning revenue earlier.
Partners plan to deliver factories in early 2027
Nvidia named Anduril, Armada, Bull, Crusoe, DayOne, Dell Technologies, Delta, Eaton, Foxconn’s Ingrasys, GIGABYTE, HPE, Lenovo, Pegatron, Quanta Cloud Technology, Schneider Electric, Vertiv and Wistron among the companies working with it on DSX POD.
Those partners are working to deliver DSX POD AI factories worldwide in early 2027, according to Nvidia.
