Jacobs releases digital twin solution for AI data centres

0

Solution set to improve speed to revenue, energy performance and operations and maintenance for AI data centres.

Jacobs has released a data centre digital twin solution that enables developers and owners to plan, simulate and optimise gigawatt-scale artificial intelligence (AI) data centres in a virtual environment – improving speed to market, energy performance and long-term operations.

The digital twin solution, developed using the NVIDIA Omniverse DSX blueprint, combines a standardised reference design for a gigawatt-scale AI data centre with a hyper-realistic virtual environment, along with advanced simulations of compute, power and cooling systems, to serve as a real-time virtual twin of an AI data centre.

The Jacobs solution allows for secure integration of components inside and outside the data centre, from on-premises power configurations to cutting-edge simulations of indoor airflow. By integrating compute, power, cooling, water systems and site specifics, the solution provides visibility from early planning through virtual commissioning, operations and maintenance.

Jacobs chair and chief executive officer Bob Pragada said: “Advanced facilities are becoming increasingly complex and costly to build and as AI infrastructure scales to the gigawatt level, owners need greater certainty before committing capital. With our deep knowledge, global workforce and decades of delivering the world’s most advanced facilities, we’re creating solutions that help owners plan more efficiently, improve performance and achieve more secure and resilient long-term operations.”

The first module of the digital twin is available for owners and operators. Future functionality is expected to deliver value across the asset lifecycle from planning and design through commissioning and construction. Future designs of the solution aim to provide a blueprint for smaller, 250-megawatt data centres, allowing for wider industry use.

The digital twin’s modular reference design enables rapid, repeatable deployment of gigawatt-scale facilities in the physical world. The reference design and digital twin solution are key components of the NVIDIA Omniverse DSX blueprint, a comprehensive framework for creating and operating gigawatt-scale AI data centres.