Johnson Controls' Guide Delivers 50MW Return, 32% Energy Gains for 1GW AI Cooling
Johnson Controls released its second AI Factory Reference Design Guide for air-cooled chillers, detailing 1-GW scalable data center cooling architectures that eliminate cooling towers and save over 12 million gallons of water daily. The guide outlines designs returning 50 MW to AI factories and boosting annual energy efficiency by 32%.
1. New AI Factory Reference Guide Launch
Johnson Controls unveiled its second AI Factory Reference Design Guide focused on air-cooled chillers, expanding its series of global data center cooling blueprints. The guide maps a comprehensive thermal chain architecture integrating high-efficiency air-cooled YORK centrifugal chillers, fan coil walls and coolant distribution units for industrial-scale AI factories.
2. Key Technical Outcomes
The blueprint supports zero-water cooling to eliminate cooling towers, saving over 12 million gallons of water per day. It details a bifurcated loop design that returns up to 50 MW to the AI factory, delivers 32% annual energy consumption improvements, 20 MW peak power savings and 30% COP gains with 27% fewer chillers.
3. Scalability and Adaptability
Designed for data centers up to 1 GW, the guide provides sizing references for 220 MW compute clusters with recommended design temperatures and operating conditions across each thermal stage. It outlines adaptable solutions for various climates, workloads and future expansion paths to maintain efficiency and reliability.
4. Strategic Impact
This reference guide reinforces Johnson Controls' leadership in thermal management and decarbonization for mission-critical facilities. By addressing water scarcity and energy efficiency challenges in AI data centers, the company aims to drive customer adoption, support sustainability goals and boost long-term revenue growth in a rapidly expanding market segment.