Siemens, NVIDIA, and Fluence have developed a reference architecture for NVIDIA DSX Vera Rubin NVL72 AI data centers, targeting a 136 MW facility with a 100 MW IT load. This design incorporates critical infrastructure elements such as utility connections and power distribution, while emphasizing Tier III concurrent maintainability and the ability to scale in phases from tens to hundreds of megawatts. In addition, Fluence introduces battery storage to enhance resilience against grid fluctuations, aligning with the trend of integrating such solutions into data center projects to ensure operational continuity.

NVIDIA: NVIDIA develops advanced computing platforms, including GPUs and data center systems optimized for AI workloads. The company is central to the news through its DSX Vera Rubin NVL72 platform, for which the new reference architecture is designed. The partnership extends NVIDIA’s ecosystem by integrating power and storage expertise for next-generation facilities.
Fluence: Fluence specializes in battery energy storage systems and grid-edge solutions for utilities and large-scale energy users. In the reported development, Fluence contributes storage capabilities to support voltage stability, black start operations, and load management in the AI data center design. This involvement positions Fluence within emerging high-demand power infrastructure projects.
Siemens: Siemens is a global technology company focused on industrial automation, electrification, and digital infrastructure solutions for energy and manufacturing sectors. In this news, Siemens provided key components for utility connections, medium-voltage distribution, and controls in the reference architecture. The collaboration highlights Siemens’ role in enabling reliable power systems for demanding AI environments.

`json
{
“Energy Integration”: “Battery storage solutions are being incorporated into data center projects to enhance resilience against grid fluctuations and support operational continuity.”,
“Infrastructure Scaling”: “Reference designs for AI facilities emphasize phased expansion and concurrent maintainability to meet evolving compute requirements.”
}
`