Trial / Demo
A new dawn for aging infrastructure. With the Uptime Institute Global Data Center Survey 2026 indicating that around half of data centre facilities are more than twelve years old, the opportunity and urgency to modernise is clear. According to JLL, retrofit activity is rising sharply, driven by the urgent need to prepare older facilities for AI driven compute densities with liquid cooling strategies seen as the only viable, scalable path to thermal resilience. At the same time, power realities are also starting to bite. Electricity demand from data centres is putting renewed emphasis on efficiency, demand shaping, and heat recovery opportunities in retrofit roadmaps.
The numbers add up
The financial case for retrofits is compelling too, with some recent studies indicating that data centre retrofits can reduce CapEx by as much as 20-40% and accelerate time-to-revenue by 6-12 months – all while leveraging existing infrastructure and grid access. In this context, data centre operators need to look to more advanced modelling methodologies to identify viable upgrades, validate efficient cooling measures, and confidently modernise legacy infrastructure.
The way forward
A combined approach integrating dynamic simulation and real operational data from the facility (otherwise referred to as a ‘digital twin’) offers a compelling pathway. The process typically begins with the development of a calibrated baseline model of the existing facility. Drawing on real building data, including metering, individual server rack data, BMS logs, and operational parameters, the model is then tuned to accurately reflect the facility’s current performance. Such a model can be used with certainty to virtually test and compare any number of retrofit interventions, for example different liquid cooling or hybrid strategies (rear door heat exchangers, direct to chip loops, liquid to liquid schemes), inclusion of free cooling, or onsite renewable/distributed energy solutions for power security. Robust simulation data, such as that produced by a Digital Twin, can help build risk-aware business cases, de-risking capital spend while also ensuing uptime and resilience.
Every scenario covered
‘What if’ scenario assessments can answer key questions such as:
• AI/GPU growth scenarios: step changes in ITE power, rack density, redundancy, and containment; evaluation of thermal compliance under peak and transient loads.
• Cooling transition planning: stage the move to liquid based cooling from rear door heat exchangers and sidecar/in row CDUs to direct to chip loops validating facility water system readiness and CDU demarcation.
• Climate and water risk: stress test hotter summers, temperature constraints, evolving wet bulb profiles, and water availability to protect WUE and uptime.
• Controls optimisation: simulate supply temperature lifts, pressure strategies, sequencing and free cooling regimes to improve resilience and efficiency.
By delivering more detailed and accurate performance insights, this approach also makes it easier to validate vendor claims surrounding new technology deployments.
To learn more about dynamic simulation and how it addresses the key challenges facing modern data centres, download the full IES whitepaper: De-risking High-Performance Data Centres with Dynamic Simulation
Discover how whole-facility modelling is transforming data centre planning, design, retrofit, and operation in an era of AI-driven infrastructure demands.
