Vertiv has announced new progress in its collaboration with NVIDIA, marking a major step toward delivering high-efficiency 800 VDC power systems designed for next-generation AI factories. The development follows their strategic alignment in May 2025, as Vertiv moves from conceptual frameworks to engineering-ready designs to support NVIDIA’s forthcoming Rubin Ultra platforms slated for 2027.
The initiative comes as data centers face unprecedented power and efficiency challenges. Traditional 54 VDC rack distribution suitable for kilowatt-scale systems falls short of meeting the megawatt-scale power needs of accelerated computing. To bridge this gap, Vertiv and NVIDIA are co-developing scalable 800 VDC architectures that integrate energy storage and optimise power delivery for large-scale AI and high-performance computing workloads.
“Larger AI workloads are reshaping every aspect of data centre design.” Said Scott Armul, executive vice president, global portfolio and business units at Vertiv. “Our development of the Vertiv 800 VDC platform translates deep systems expertise into the scalable, efficient power foundations that AI factories demand.”
“Powering the next generation of megawatt-scale AI factories requires a fundamental shift in power architectures. Our collaboration with Vertiv is key to building that foundation.” Added Dion Harris, senior director for HPC, Cloud and AI Infrastructure at NVIDIA.
Vertiv’s 800 VDC platform builds on decades of experience in DC power for telecom and industrial environments. The company is already working with early-stage AI factory projects. Its reference architecture forms the design basis for gigawatt-scale deployment. Supported by over 4,000 field engineers. Vertiv’s global service capability underpins safe, reliable operation of these high-density environments. This positions the company as a critical infrastructure partner in the emerging AI era.
