Johnson Controls, a global leader in smart, healthy, and sustainable buildings, has launched its groundbreaking Reference Design Guide Series for 1 Gigawatt (GW) AI data centers. These comprehensive guides meticulously map the entire thermal chain, presenting sophisticated cooling architectures precisely tailored for diverse compute densities, geographical locations, and elevations. The series kicks off with an in-depth blueprint for water-cooled chiller plants, with future installments set to explore air-cooled and absorption chiller solutions.
As artificial intelligence continues its transformative march across industries, the sheer scale and complexity of data center infrastructure are evolving at an unprecedented pace. The ability to efficiently manage thermal loads at the gigawatt scale has become a critical enabler for AI innovation. The industry faces mounting pressure to deliver facilities that are not only high-performing but also inherently sustainable and future-ready. Johnson Controls’ Reference Design Guide Series directly addresses this challenge, outlining strategies to achieve industry-leading energy and water efficiency (PUE and WUE) while maintaining the crucial flexibility to scale across a multitude of climates and operational requirements.
The guide details a complete thermal architecture that robustly supports both liquid and air-cooled IT loads. This is achieved through the seamless integration of computer room air handlers (CRAHs), innovative fan coil walls, coolant distribution units (CDUs), and high-efficiency YORK centrifugal chillers. It provides precise sizing guidance for 220MW compute quadrants and meticulously defines temperature and operating conditions across all major facility loops, including the critical Technology Cooling System (TCS) loops designed to support next-generation GPUs.
Key outcomes enabled by this updated design include:
- Zero Water Consumption: An innovative, fully water-free heat rejection process utilizing dry coolers, significantly reducing operational costs and advancing critical sustainability objectives.
- Future-Ready Thermal Flexibility: High-temperature TCS loop readiness ensures seamless compatibility with forthcoming GPU architectures, safeguarding future investments.
- Optimized High-Density AI Performance: Direct alignment with NVIDIA DSX reference architecture facilitates the scalable deployment of 1-GW-class AI Factories.
- Energy-Efficient Operation: Elevated condenser water temperatures, bifurcated loops, and high-lift YORK chillers combine to deliver industry-leading Power Usage Effectiveness (PUE) and significantly improved annualized efficiency.
“AI Factories are production facilities — the places where intelligence is manufactured at an industrial scale,” remarked Austin Domenici, vice president & general manager, Johnson Controls Global Data Center Solutions. “By supporting the NVIDIA DSX reference architecture and enhancing water and energy efficiency in the cooling process while maintaining high-temperature loop compatibility, our Reference Design Guide empowers customers to deploy gigawatt-scale AI infrastructure that is scalable, repeatable, resilient, and, critically, sustainable.”