Balancing Data Center Computing Demands and Energy Efficiency

Most people don’t think about the amount of energy they use every day or how the computing required for their daily social media check-ins, generative AI searches and online shopping sprees impacts that number. The computation of processors in a data center makes all those things – and more – happen. Data centers also handle the processing needs of cloud computing, the Internet of Things, remote connectivity and the automation of businesses, organizations and governments of all sizes. The proliferation of smart, connected devices – and new technologies – is driving the development of new data centers and updates to existing facilities globally. It’s also increasing global data center capacity, which is expected to grow from 59 GW in 2025 to 122 GW by the end of 2030. ⁱ

This requires energy. Lots of energy. It's estimated that data centers could account for up to 10% of total electricity demand growth globally by 2030ⁱⁱ. This begs the question: how can data centers use energy more efficiently?

The answer may lie in integrating new systems and capabilities with a deliberate focus on better managing energy use. Below are three areas for data center operators to consider when evaluating their site’s capabilities.

• Upgrading IT and OT hardware. Replacing existing IT and OT hardware with the latest versions often offers improvements in energy efficiency due to advancements in chip design, power management and power-saving features such as dynamic frequency scaling, and even product designs that support better airflow. It’s also important to consider how to integrate data from both IT and OT systems within a data center using digital control systems and automation. This integration can benefit from analytical tools that can optimize energy use, asset utilization and power source carbon footprint as well as capabilities like predictive maintenance for early identification and remediation of inefficient equipment that may be using more energy than needed.
• Evaluating cooling techniques. Your cooling systems may benefit from deployment of solutions like thermal energy storage (TES) to deliver facility-level peak shaving or rack-level emergency burst cooling. When using a TES at the facility level, the data center can reduce the overall energy consumed as well as the associated carbon footprint. It’s also important to consider the refrigerants used in your facility. Replacing legacy, high global warming refrigerants in CRAC and DX cooling systems with low global warming potential refrigerants can help reduce a data center’s carbon footprint as well.
• Deploying energy storage capabilities. Adding battery energy storage systems (BESS) to your data center can help solve several challenges. It can store variable renewable energy, support firmness of supply, meet backup power requirements and allow operators to leverage opportunities for daily price arbitrage while avoiding high carbon-peak grid power costs. Overall, it can help make your facility more energy resilient to help enable your uptime goals.

These actions – when taken individually or collectively – may help to better manage energy use in a data center. They can also be used when planning and specifying new data centers.

To learn more about how Honeywell can help you improve the energy use of your data center, please read our whitepaper, Design for More Efficient Data Centers.

ⁱGoldman Sachs, "AI to drive 165% increase in data center power demand by 2030," February 4, 2025. [Accessed April 1, 2025]
ⁱⁱ The Oxford Institute for Energy Studies, "Global electricity demand: what's driving growth and why it matters?" Rahmatallah Poudineh, January 2025. [Accessed April 1, 2025]