Siemens has unveiled a new reference design for NVIDIA DSX Vera Rubin NVL72 AI factories, as the industry races to meet the escalating infrastructure demands of next-gen artificial intelligence workloads.
It was developed in collaboration with NVIDIA and Fluence and incorporating nVent-aligned design considerations.
The blueprint translates NVIDIA's AI factory vision into a deployable, industrialized electrical, power and controls architecture aimed at hyperscalers, colocation providers and specialized cloud infrastructure operators.
The launch comes as AI factories rapidly reshape data center design, with increasingly power-hungry platforms forcing operators to rethink everything from site selection and grid connectivity to capital deployment and speed-to-market.
Built for a total facility capacity of 136 MW and an IT load of 100 MW, Siemens' reference architecture delivers an end-to-end electrical and controls framework extending from the utility connection through medium-voltage distribution and modular low-voltage power blocks to the rack level.
The design is based on a Tier III concurrent maintainability model, allowing any single component to be serviced without disrupting IT operations.
The modular architecture is designed to scale alongside demand, enabling operators to expand from initial deployments of tens of megawatts to facilities exceeding hundreds of megawatts without requiring a fundamental redesign.
The reference design also incorporates nVent-aligned electrical parameters to ensure compatibility with NVIDIA workloads and system architectures. Siemens said a future supplement will add advanced thermal management capabilities as liquid cooling becomes increasingly critical for high-density AI environments.
“nVent has deployed more than two gigawatts of liquid cooling capacity globally,” said Sara Zawoyski, President, nVent Systems Protection.
“That operational experience is what allows us to help partners like Siemens translate reference architectures into deployable thermal solutions that perform reliably from day one at this scale. Platforms like NVIDIA Vera Rubin NVL72 are pushing rack densities well beyond what traditional air-cooled infrastructure can support.”
The DSX Vera Rubin reference design is intended to help AI factory operators deploy extreme-density AI infrastructure more quickly while reducing risk and improving operational predictability.
Siemens said the architecture supports DSX MaxLPS, enabling operators to maximize computing output and token production within fixed power constraints while shortening deployment schedules.
“Siemens’ deep expertise in power systems and controls engineering, modular infrastructure, protection, and industrialized delivery is really evident in this latest joint reference architecture design,” said Ruth Gratzke, President of Siemens Smart Infrastructure USA.
“Our pre-engineered, prefabricated, and factory-tested medium- and low-voltage skids help minimize on-site construction complexity, shorten commissioning cycles, and improve quality, safety, and repeatability across deployments. Further, our automation and digital twin strategies deployed in this reference help ensure that facilities are brought online faster and with greater potential to produce tokens at scale.”
Energy storage also plays a central role in the architecture. Fluence's battery energy storage technology is integrated into the design to provide operational flexibility and resilience in power-constrained environments, helping AI factories maintain reliability while scaling rapidly.
“Our Smartstack platform is central to this new architecture, transforming the grid into an accelerator for compute,” said Jeff Monday, Fluence Chief Growth Officer.
“By providing essential capabilities like voltage and frequency ride through, black start, grid demand response, and AI load smoothing, we are enabling our customers to build the AI factories of the future faster and more reliably.”
The design further integrates a centralized Integrated Data Center Management Suite, providing operators with a unified view across power, cooling and compute infrastructure.
The announcement underscores the growing convergence of power infrastructure, energy storage, cooling technology and AI computing as the industry prepares for the next generation of ultra-dense AI factories.
With power availability, deployment speed and operational efficiency emerging as critical competitive advantages, vendors are increasingly packaging complete infrastructure blueprints designed to accelerate large-scale AI deployments.
