This technical paper outlines a reference blueprint for a 100 MW hyperscale AI data center designed to support NVIDIA GB200 NVL72-based systems like NVIDIA DGX GB200 or systems from NVIDIA partners. It presents a Tier III capable architecture that integrates Siemens’ industrial-grade electrical systems with NVIDIA GB200 NVL72 systems and nVent liquid cooling technology in the UL market.
The paper covers pod-level modularity, power distribution, cooling, control and automation, and operational scalability. It also provides definitions, terminology, schematics, and an appendix to support technical planning, design coordination, and stakeholder alignment.