ETAP and Schneider Electrical have collectively launched what they’re calling the world’s first digital twin able to simulating each side of an AI Manufacturing unit’s energy utilization, from the grid proper all the way down to chip stage.
Unveiled at NVIDIA’s GPU Expertise Convention, the answer is ready to leverage NVIDIA Omniverse know-how to assist information centre and AI operators higher handle the rising calls for of their infrastructure.
How the digital twin works
It’s no secret that AI workloads proceed to put a big burden on information centres, with superior coaching and inference processes demanding excessive rack energy densities. In reality, studies recommend that AI coaching duties use 10-15 instances extra energy than conventional information centres had been designed to deal with.
Meaning it’s not at all times instantly apparent to operators what they want for his or her AI-ready information centres, nor how they’ll successfully handle that rising demand for energy. That’s the place the brand new digital twin from ETAP and Schneider Electrical is available in.
By integrating ETAP’s superior energy system design experience with NVIDIA Omniverse, the brand new digital twin presents real-time modelling, ‘what-if’ eventualities, and predictive upkeep methods that declare to ship higher reliability, sustainability, and price financial savings.
Crucially, this collaboration introduces a ‘Grid to Chip’ strategy – aiming to grasp and optimise energy consumption not solely on the rack but in addition on the chip stage. With AI workloads ever-growing, exact management over load behaviour may show very important to operators’ backside strains.
“As AI workloads develop in complexity and scale, exact energy administration is important to making sure effectivity, reliability, and sustainability,” mentioned Dion Harris, Senior Director of HPC and AI Manufacturing unit Options at NVIDIA.
“Via our collaboration with ETAP and Schneider Electrical, we’re providing information centre operators unprecedented visibility and management over energy dynamics, empowering them to optimise their infrastructure and speed up AI adoption whereas enhancing operational resilience.”
‘AI Manufacturing unit of tomorrow’
By opening the door to enhanced system planning, dynamic energy circulate evaluation, and predictive upkeep, ETAP and Schneider Electrical consider information centres of the longer term won’t solely reduce operational prices but in addition bolster effectivity, reliability and sustainability, paving the best way for what they name the ‘AI Manufacturing unit of tomorrow’.
“This collaboration represents greater than only a technological answer,” mentioned Tanuj Khandelwal, CEO of ETAP.
“We’re essentially reimagining how information centres could be designed, managed, and optimised within the AI period. By bridging electrical engineering with superior virtualisation and AI applied sciences, we’re creating a brand new paradigm for infrastructure administration.”
Pankaj Sharma, Government Vice President for Knowledge Facilities, Networks & Providers at Schneider Electrical added, “Collaboration, pace, and innovation are the driving forces behind the digital infrastructure transformation that’s required to accommodate AI workloads. Collectively, ETAP, Schneider Electrical, and NVIDIA aren’t simply advancing information centre know-how – we’re empowering companies to optimise operations and seamlessly navigate the ability necessities of AI.”
