ETAP and Schneider Electric introduce the first Digital Twin that simulates the energy needs of AI factories, together with NVIDIA Omniverse.

• The Digital Twin provides better insights and greater control over the electrical systems and power requirements of AI factories. 
• The collaboration integrates ETAP's advanced Electrical Digital Twin technology with NVIDIA Omniverse™ cloud APIs. 
• Operators benefit from advanced energy efficiency, predictive maintenance, and lower costs. 

Schneider Electric, leader in the digital transformation of energy management and automation, and ETAP, industry and technology leader in the design and operation of energy systems, are introducing a Digital Twin that can accurately model and simulate the energy needs of AI factories. Schneider Electric and ETAP are using the NVIDIA Omniverse™ Blueprint for this purpose. Data from mechanical, thermal, network and electrical systems are combined to simulate the operation of an AI factory. The collaboration aims to transform the design and operation of AI factories by providing greater insight and control over electrical systems and power requirements. This offers the potential for significant improvements in efficiency, reliability and sustainability.

Previously, only simple visualisations of electrical systems were possible. Thanks to the integration of ETAP and NVIDIA Omniverse-technologies, an advanced AI Factory Digital Twin can be created, in which multiple dynamic systems work together seamlessly. ETAP's modelling technology creates a virtual replica of a data centre's electrical infrastructure and enriches it with real-time data about the electricity system, advanced analyses and valuable insights. Intelligent algorithms analyse and predict power consumption and distribution patterns, providing unique insights into:

• Advanced design and simulation of electrical systems
• Dynamic “What if” scenario analysis
• Real-time performance monitoring of the electrical infrastructure
• Advanced optimisation of energy efficiency
• Predictive maintenance and system reliability assessment
• Infrastructure requirements based on power consumption, which can contribute to lower total cost of ownership

AI workloads are causing a significant increase in data centre power consumption, from large-scale training clusters to edge inference servers. Unlike traditional computing tasks, AI tasks – especially model training and complex inference processes – require enormous computing power. This results in higher power density per rack. As AI becomes more widely used, start-ups, enterprises, colocation providers and internet giants must rethink the design and management of data centres to meet the growing need for energy efficiency.

The collaboration between ETAP and NVIDIA introduces an innovative “Grid to Chip” approach that addresses the challenges of energy management, performance optimisation and energy efficiency in the age of AI. Currently, data centre managers can only estimate average power consumption at the rack level. ETAP's new digital twin focuses on more accurate modelling of dynamic loads at the chip level. This improves the design of power systems and optimises energy efficiency.

This collaboration highlights ETAP and NVIDIA's commitment to driving innovation in the data centre industry. It enables companies to optimise their operations and effectively manage the challenges associated with AI workloads. The collaboration aims to improve data centre efficiency while enhancing the reliability and performance of the power grid.

“As AI workloads become more complex and large-scale, precise energy management is critical to ensuring efficiency, reliability and sustainability,” said Dion Harris, Senior Director of HPC and AI Factory Solutions at NVIDIA. “Through our collaboration with ETAP and Schneider Electric, data centre managers gain unparalleled insight and control over energy dynamics. This enables them to optimise their infrastructure, accelerate AI deployment and strengthen operational resilience at the same time.”
​

“This collaboration represents more than just a technological solution,” says Tanuj Khandelwal, CEO of ETAP. “We are fundamentally rethinking how data centres are designed, managed and optimised in the AI era. By connecting electrical engineering with advanced virtualisation and AI technologies, we are creating a new paradigm for infrastructure management.”

Pankaj Sharma, Executive Vice President for Data Centres, Networks & Services at Schneider Electric adds: “Collaboration, speed and innovation are at the heart of the digital infrastructure transformation needed to support AI workloads. ETAP, Schneider Electric and NVIDIA are going beyond simply improving data centre technology – they are enabling businesses to optimise their operations and effortlessly navigate the power requirements of AI.”