Foxconn subsidiary 鴻佰科技 has become a global partner for NVIDIA’s DSX POD program, linking the company to a modular AI data center design based on the Vera Rubin computing platform.
DSX POD is a blueprint for partner-built facilities, with the design meant to support modular AI data center deployments. Under the approach, partners prepare racks at manufacturing sites, wire them for power, install liquid-cooling components and complete system checks before shipment.
The factory work is intended to reduce time at customer locations. In its earlier deployment outline for prepared sites, NVIDIA estimated an on-site rollout of 14 days when power, cooling and network service are ready; it put a conventional deployment at six months for a comparable stage.
An NVL72 rack combines 72 Rubin GPUs with 36 Vera CPUs. NVIDIA previously named Dell Technologies, HPE and Lenovo among companies working with the DSX POD design.
Worldwide deliveries by partners are expected to begin in early 2027.
