Home / About / Thought Leadership / Case Studies / From Cooling Challenges to Climate Solutions
AI is transforming data centers by pushing energy and cooling systems to their limits, driving infrastructure to evolve to handle high-density demands while unlocking opportunities to capture waste heat for thermal energy networks.
Civil Engineer
sransbottom@wendelcompanies.com
1.877.293.6335
The use of artificial intelligence is transforming data centers. As organizations race to adopt AI technologies, from large language models to real-time analytics, the infrastructure behind the scenes is evolving rapidly. Today’s AI data centers increasingly resemble high-performance supercomputing facilities, packed with dense arrays of Graphics Processing Units (GPUs) and specialized hardware designed to process vast amounts of data and computation.
This evolution comes with a steep energy price. Traditional data centers typically consume 5 to 10 kilowatts of power per rack. In contrast, AI-driven data centers often require as much as 200 kilowatts, sometimes in a much smaller physical footprint. This level of energy intensity places unprecedented pressure on cooling systems, power distribution, and resiliency planning. However, it also presents an opportunity to utilize data centers as part of a sustainable thermal network.
AI Data centers are being forced to rethink their approach to cooling. Traditional air-based systems were sufficient for past workloads but no longer meet the demands of high-density AI equipment. Even with advanced liquid cooling methods, only about 80% of the heat generated by data centers is removed. Air-based systems must still manage the remaining heat, resulting in complex hybrid cooling environments that require careful engineering. At maximum efficiency, these systems transfer heat out of the data center through the thermal loop, enabling reliable operations while allowing the recaptured heat to be reused. Increasingly, data centers are being integrated into thermal energy networks, where their excess heat is used to warm nearby buildings. These systems improve overall energy efficiency and reduce environmental impact by treating heat as a resource rather than a waste product.
Power reliability becomes a mission-critical concern as AI workloads demand more consistent and dense energy delivery. Even a brief disruption in cooling or electricity can damage sensitive systems or halt computation, making backup infrastructure—such as batteries, generators, and automated recovery protocols—more crucial than ever.
Energy storage planning must evolve in parallel. As we rethink data center infrastructure, we must also evaluate how clean and decarbonized our energy sources are and whether our backup strategies align with broader sustainability goals. Moving away from fossil-fuel-based generators and toward renewables or low-carbon alternatives will be key to building resilient yet environmentally responsible systems.
Thermal loops, increasingly used to carry heat away from data centers, enable more efficient heat recovery and reduce reliance on high-powered air-cooled systems. By limiting the need for large, constantly running fans, these systems help reduce noise pollution, a growing concern as data centers are sited closer to communities.
Rather than fully committing to AI, many businesses are opting for a hybrid approach, managing some workloads on-site, while others are processed in the cloud or through colocation providers. This reinforces the need for customized design solutions tailored to real operational needs, rather than one-size-fits-all infrastructure.
AI fundamentally changes data centers’ requirements regarding energy, cooling, and architecture. But this shift is not just a challenge; it’s a chance to build smarter, more efficient, and more sustainable systems from the ground up. As data centers become both the engines of AI and potential assets in broader energy ecosystems, we have an opportunity to redefine their role, not just as power consumers, but as integrated components of sustainable urban infrastructure. Our decisions today will shape the digital and environmental legacy of the AI era.