Flex Unveils NVIDIA AI Modules Cutting Deployment Time 30%
Flex's new NVIDIA Omniverse DSX reference designs integrate 800 VDC power racks with advanced liquid cooling and high-density IT racks to enable 30% faster AI factory deployment. These prefab modular systems leverage Flex’s 18M sq ft footprint across 35+ American sites to scale and accelerate U.S. AI infrastructure rollouts.
1. AI Factory Reference Designs
Flex introduced pre-engineered modular systems under its AI Infrastructure Platform, co-developed with NVIDIA for Omniverse DSX Blueprint. These modules combine power, high-density liquid-cooled IT racks and advanced networking into factory-built units, with an 800 VDC Power Rack disaggregating power components to maximize GPU density for the NVIDIA Vera Rubin platform.
2. Deployment Efficiency Gains
The reference designs support phased migration from traditional AC setups to 800 VDC architectures, cutting on-site complexity and accelerating AI factory builds by up to 30% compared with conventional construction methods.
3. Vast Americas Footprint
Flex leverages more than 18 million square feet of manufacturing space across 35+ locations in the Americas—including a 400,000 sq ft data center infrastructure facility in Dallas—to shorten lead times and serve U.S. data center operators with rapid AI infrastructure deployments.