Etap and Schneider Electric, together with Nvidia, are launching a digital twin that simulates the energy consumption of AI data centers. The technology aims to provide data center managers with more control and insight into power consumption.
Digital Twin for Electrical Infrastructure
The new digital twin combines Etap’s electrical modeling technology with the visualization and simulation power of NVIDIA Omniverse. The solution enables virtual modeling of the entire electrical system of an AI data center, including real-time data on energy consumption, performance monitoring, and system behavior. This allows data centers to simulate ‘what-if’ scenarios, optimize energy efficiency, and perform predictive maintenance, among other things.
AI applications such as model training and inference require significant computing power. This results in higher power density in racks and increases pressure on the power infrastructure. The digital twin aims to provide a solution to this challenge by accurately predicting power consumption at the chip level instead of just at the rack level.
Greater Control Over AI Energy Requirements
With this collaboration, the three companies introduce a so-called ‘Grid to Chip’ approach. This approach provides insight into the entire journey from energy generation to energy consumption by AI processes. The technology is aimed at companies that are rolling out AI applications while wanting to optimize their energy management. Data centers can design and manage their infrastructure more efficiently, potentially leading to lower costs and increased reliability.
The companies involved emphasize that this digital twin is a tool for organizations looking to adapt their data center design to the increasing energy demands of AI. The collaboration focuses on various sectors, from startups to colocation providers and larger enterprises.
read also