Vertiv (NYSE: VRT), a global provider of critical digital infrastructure and continuity solutions, has announced plans to introduce a comprehensive 800 VDC power portfolio in alignment with NVIDIA’s roadmap for next-generation AI-centric data centers. The new offering is slated for release in the second half of 2026, positioning Vertiv to support upcoming NVIDIA Kyber and Rubin Ultra platforms.
The strategic alignment is aimed at helping data center operators and hyperscalers future-proof their infrastructure to meet growing demands in artificial intelligence (AI) workloads. According to Vertiv, the shift toward 800-volt direct current (VDC) architectures offers a more efficient method of centralized power delivery, particularly as rack power densities in AI environments begin to exceed 300 kilowatts.
“As GPUs evolve to support increasingly complex AI applications at giga-watt scale, power and cooling providers need to be equally innovative,” said Scott Armul, Executive Vice President of Global Portfolio and Business Units at Vertiv. “While the 800 VDC portfolio is new, DC power isn’t a new direction for us—it’s a continuation of what we’ve already done at scale.”
Vertiv’s upcoming 800 VDC lineup will include centralized rectifiers, high-efficiency DC busways, rack-level DC-DC converters, and backup power systems compatible with DC architecture. These will complement Vertiv’s existing portfolio of alternating current (AC) power systems, offering data centers increased flexibility in power design.
The company noted that its experience in high-voltage DC systems spans more than two decades, including extensive deployments of ±400 VDC infrastructure for global telecom and industrial applications. Vertiv has also expanded its expertise in this area through targeted acquisitions in the early 2000s.
The move toward 800 VDC systems aligns with a broader industry trend to enhance energy efficiency and reduce thermal losses in data centers supporting high-performance AI applications. Vertiv’s designs aim to minimize copper use, lower current levels, and reduce overall power losses—all key considerations in hyperscale environments.
NVIDIA’s roadmap emphasizes the importance of scalable, efficient infrastructure for AI data centers. By aligning with this direction, Vertiv aims to ensure its infrastructure solutions remain one generation ahead of GPU advancements, allowing customers to synchronize their deployments with evolving compute needs.
Vertiv’s new power architecture will also serve as a critical component of its “unit of compute” strategy, which focuses on modular, interoperable infrastructure tailored for homogeneous AI zones. This systems-level approach is designed to integrate power, cooling, and other core components into a scalable, unified solution.
