NVIDIA unveils modular DSX POD infrastructure to speed deployment

NVIDIA introduced DSX POD at GTC Berlin, a reference design for complete Vera Rubin AI factories assembled and validated inside modular containers before delivery. The company says the approach can reduce deployment time at a prepared site from six months to as little as 14 days. Units arrive ready to connect to existing power, cooling, networking and supporting infrastructure.

The design moves computing and networking integration, cabling, rack-level power and liquid cooling installation into manufacturing facilities. Partners complete wiring, piping and functional testing before shipment while site preparation proceeds separately. NVIDIA says this reduces exposure to component delays and complex onsite work, allowing operators to begin model training and production inference sooner. The intended economic benefit is an earlier start to serving customers and generating revenue.

Worldwide delivery is planned for early 2027 through a computing and infrastructure partner ecosystem that includes Bull, Dell Technologies, HPE, Lenovo, Schneider Electric and Vertiv. DSX POD builds on NVIDIA’s broader DSX infrastructure platform with a distinct containerized architecture designed for repeatable manufacturing and deployment at scale.