nexttech.be
en
The knowledge platform for technology, innovation and digital transformation
Schneider Electric, a global leader in energy technology, together with NVIDIA and industrial software company AVEVA, has announced new developments to design, simulate, build and manage next-generation AI data centre infrastructure. The announcements were made during NVIDIA GTC in San Jose. These include a new NVIDIA Vera Rubin reference design that validates the power requirements and cooling of the latest NVIDIA rack-scale architectures, the integration of advanced digital-twin capabilities into the NVIDIA Omniverse DSX Blueprint and its ecosystem, and early tests with agentic AI for alarm management in data centres, based on NVIDIA Nemotron-models.
With these announcements, Schneider Electric and NVIDIA are strengthening their existing partnership and laying a solid foundation for the development of gigawatt AI factories designed for maximum efficiency.
The new AI reference design is one of the first designs for the NVIDIA Vera Rubin NVL72-racks. The validated reference design covers both power supply and cooling and is integrated with the reference designs for control systems From Schneider Electric.
The design addresses key infrastructure requirements for NVIDIA's latest rack-scale systems:
The reference design was validated using ETAP models for electrical system design and ITD CFD models for layout and airflow optimisation.
Furthermore, AVEVA, global supplier of industrial software and part of Schneider Electric, has teamed up with NVIDIA to launch a new lifecycle digital-twin architecture announced. It increases the efficiency of GPUs and accelerates the rollout of AI factories on a large scale.
Schneider Electric is developing SimReady kits and digital twins for this purpose within NVIDIA Omniverse, supported with AVEVA's software. The announcement integrates AVEVA's engineering and operations software into the NVIDIA Omniverse DSX Blueprint ecosystem. The technology shortens time-to-market by cleverly combining simulations, digital visualisation and design tools to streamline the engineering process.
Once the system architecture is built in the NVIDIA Omniverse environment, AVEVA performs simulations in multiple domains to validate operational behaviour under realistic conditions. This involves the use of computational models for power distribution, thermodynamics, air flows and control engineering, among others. These simulations allow designs to be iteratively optimised, different scenarios to be quickly tested and the system to be fully validated before the physical environment is built. This shortens engineering cycles and increases the accuracy of implementations.
“As AI workloads increase in both size and complexity, the margin for error in data centre design becomes extremely small,” said Manish Kumar, executive vice president of Secure Power & Data Centres at Schneider Electric. “Large-scale AI requires tight integration of electrical, cooling and digital architectures that support extreme performance requirements and maximum energy efficiency. By combining advanced software, digital twins and validated reference designs, operators can simulate and optimise their infrastructure before a single rack is installed. This reduces risk, accelerates deployment and ensures the efficiency and resilience needed for the next generation of AI factories.”
“Gigawatt-scale AI factories require a fundamentally new kind of energy-efficient and predictable infrastructure,” said Vladimir Troy, Vice President AI Infrastructure at NVIDIA. Together, NVIDIA and Schneider Electric provide the power, cooling and digital-twin architectures needed to accelerate time-to-market for customers worldwide.”
Schneider Electric also announced at the event that it has successfully tested the NVIDIA Nemotron model for a new agentic-AI application for alarm management. This development responds to a long-standing challenge in the data centre industry: interpreting alarms at the system level to identify the root cause and determine appropriate action. Using real-time streaming IoT data from different systems, Schneider Electric's agentic-AI solution can independently analyse problems, diagnose and recommend actions through integrated tools. In collaboration with expert technicians, this ensures faster and more consistent problem resolution, fewer unnecessary interventions and increased operational resilience. The move underlines Schneider Electric's ambition to further innovate asset performance management with AI.
The announcements build on a long history of innovation between Schneider Electric and NVIDIA:
Neem dan rechtstreeks contact op met Schneider Electric.
Contact opnemen
"We provide digital energy management and automation solutions for efficiency and sustainability. We enable everyone to make the best use of their energy and resources, so that 'Life Is On' applies to everyone, anytime, anywhere."
Bekijk Bedrijfsprofiel
