Vertiv aligns with NVIDIA's AI roadmap to stay one GPU generation ahead, enabling customers to deploy their power and cooling infrastructures in sync with NVIDIA's next-generation computing platforms. Vertiv will provide a complete and integrated power, cooling, and services infrastructure to support AI factories and other data center deployments.
As the power requirements of racks in AI environments continue to scale beyond 300 kW, the 800 VDC architecture offers a more efficient and centralized power supply by reducing copper usage, current, and heat losses. Vertiv's future portfolio will include centralized rectifiers, high-efficiency DC power bars, rack-level DC-DC converters, and DC-compatible backup systems, expanding its extensive power management portfolio, which already includes robust AC power solutions.
“As GPUs evolve to support increasingly complex AI applications at gigawatt scale, power and cooling providers need to be equally innovative to deliver efficient, high-density solutions for AI factories. While the 800VDC portfolio is new, DC power is not new to us—it’s a continuation of what we’ve already done at scale,” said Scott Armul, Executive Vice President of Global Portfolio and Business Units at Vertiv. “We’ve spent decades deploying high-voltage DC architectures in global telecom applications, industrial environments, and data centers. We’re preparing for this transition from a position of strength and real-world experience to meet the demands of AI factories.”.
Vertiv's DC power expertise spans more than two decades of ±400 VDC deployments, further enhanced by strategic acquisitions in the early 2000s. These solutions support critical loads in global telecommunications networks, integrated microgrids, and mission-critical facilities. This foundation establishes Vertiv as a trusted leader in the safe design, deployment, and operation of high-voltage DC architectures, with a reliable portfolio, scale, and long-term functionality.
Designed for homogeneous AI zones in hyperscale environments, Vertiv's 800 VDC portfolio is a key pillar of its "compute unit" strategy—a system-level design created to enable all infrastructure components to interoperate as a modular and scalable system, meeting the infrastructure demands of the next generation of GPUs. Vertiv's operations with both AC and DC architectures represent a strategic differentiator in the evolving landscape of AI data centers.
