Schneider Electric Launches AI Data Center Reference Designs Co-Engineered with NVIDIA
Schneider Electric, a global leader in energy management and automation, today unveiled a set of innovative reference designs developed in collaboration with NVIDIA. These designs are crafted to accelerate deployment timelines and assist data center operators in adopting AI-ready infrastructure solutions that meet the demands of modern, high-performance computing. By combining integrated power management with advanced liquid cooling technologies, Schneider Electric’s new reference designs provide a robust foundation for AI infrastructure that is both scalable and efficient.
The first of the newly introduced reference designs offers the industry’s only comprehensive framework for integrated power management and liquid cooling control systems. This design incorporates Schneider Electric’s Motivair liquid cooling technology, enabling operators to seamlessly manage complex AI infrastructure components. Beyond managing physical infrastructure, the control system design is fully interoperable with NVIDIA Mission Control, NVIDIA’s AI operations and orchestration platform. This integration allows for advanced cluster and workload management, helping operators optimize AI workloads in real time. The reference design is also compatible with Schneider Electric’s NVIDIA Grace Blackwell systems data center designs, ensuring operators can adopt the latest advancements in accelerated computing while maintaining full oversight of power and cooling systems.
The second reference design is tailored for AI factory deployments of up to 142 kW per rack, specifically supporting NVIDIA GB300 NVL72 racks within a single data hall. Designed to serve as a blueprint for the next-generation NVIDIA Blackwell Ultra architecture, this design includes guidance across four critical technical domains: facility power, facility cooling, IT space, and lifecycle software. To meet global standards, the reference design is available in configurations adhering to both ANSI (American National Standards Institute) and IEC (International Electrotechnical Commission) specifications, providing flexibility for operators worldwide.
Engineered for Speed and Scalability
In the fast-moving world of AI, data center operators face increasing pressure to deploy high-density, GPU-accelerated clusters quickly and efficiently. Reference design frameworks like those offered by Schneider Electric are critical in overcoming deployment challenges. By providing validated and thoroughly documented physical infrastructure designs, Schneider Electric enables operators to implement next-generation power and liquid cooling controls before the arrival of cutting-edge AI hardware. This approach allows operators to optimize costs, efficiency, and reliability while staying ahead of technological advancements. Schneider Electric’s fully engineered reference designs essentially lay the groundwork for AI factories of the future, empowering operators to meet surging AI demand proactively.
“Schneider Electric is streamlining the process of designing, deploying, and operating advanced AI infrastructure with its new reference designs,” said Jim Simonelli, Senior Vice President and Chief Technology Officer at Schneider Electric. “Our latest reference designs, featuring integrated power management and liquid cooling controls, are future-ready, scalable, and co-engineered with NVIDIA for real-world applications. These solutions enable data center operators to keep pace with the rapidly increasing demand for AI while ensuring reliability and efficiency.”
Scott Wallace, Director of Data Center Engineering at NVIDIA, highlighted the transformative potential of the collaboration: “We are entering a new era of accelerated computing, where integrated intelligence across power, cooling, and operations will redefine data center architectures. With its latest controls reference design, Schneider Electric connects critical infrastructure data with NVIDIA Mission Control, delivering a rigorously validated blueprint that enables AI factory digital twins and empowers operators to optimize advanced accelerated computing infrastructure.”
‘Plug-and-Play’ End-to-End Controls System
At the heart of the controls reference design is a “plug-and-play” architecture based on the MQTT protocol, connecting edge devices and facility controls for energy management and liquid cooling. This approach bridges operational technology (OT) infrastructure with information technology (IT) systems, allowing operators to harness real-time data from every layer to optimize performance across AI workloads.
The controls reference design emphasizes seamless interoperability between building management systems and AI infrastructure management software. It introduces guidance for measuring AI rack power profiles and establishes redundant systems for power and cooling to ensure resiliency in both white and grey space environments. The outcome is an end-to-end controls system that optimizes uptime, reliability, and peak performance for AI deployments.
Key features of the controls reference design include:
- A standardized interface for publishing power management and liquid cooling control data for consumption by local applications, upstream systems, digital twins, AI/ML tools, and enterprise software.
- Management of redundancy across cooling and power distribution infrastructure, including coolant distribution units (CDUs) and remote power panels (RPPs).
- New methodologies for monitoring AI rack peak power and power quality, ensuring accurate measurement and efficient operation.
High-Density AI Cluster Deployment with NVIDIA GB300 NVL72
The NVIDIA GB300 NVL72 reference design supports high-density AI clusters with maximum rack power of 142 kW, such as NVIDIA DGX SuperPOD deployments with DGX GB300 systems. Data halls built using this reference design can host three GB300 NVL72 clusters powered by up to 1,152 GPUs, utilizing liquid-to-liquid CDUs and high-temperature chillers to maintain optimal operating conditions.
Schneider Electric’s reference design leverages advanced ETAP and EcoStruxure IT Design CFD models, enabling operators to create digital twins for simulating power and cooling scenarios. These simulations allow for precise optimization of infrastructure tailored to specific operational requirements. The GB300 NVL72 reference design builds on Schneider Electric’s earlier work with NVIDIA GB200 NVL72 clusters, continuing a collaborative effort to provide fully engineered, validated, and tested AI infrastructure solutions.
A Long-Term Partnership for AI-Ready Data Centers
These new reference designs reinforce Schneider Electric’s ongoing collaboration with NVIDIA, addressing the AI data center industry’s most pressing needs. In addition to the newly announced designs, Schneider Electric has developed nine other AI reference designs, covering scenarios such as prefabricated modules, retrofitted data centers, and purpose-built AI infrastructure for both NVIDIA GB200 and GB300 NVL72-based clusters.
By delivering fully engineered and validated frameworks, Schneider Electric is helping data center operators worldwide meet the demands of next-generation AI workloads with confidence. These reference designs not only provide scalable and efficient infrastructure blueprints but also pave the way for the creation of AI factories capable of adapting to rapidly evolving technology. With integrated power management, advanced liquid cooling, and seamless software interoperability, Schneider Electric and NVIDIA are setting a new standard for AI-ready data centers, ensuring operators can keep pace with the accelerating demands of artificial intelligence.
