Executive Roundtable: Cooling, Costs, and Integration in the AI Data Center Era

Brandon Peterson, CoolIT Systems:  As liquid cooling expands alongside AI, minor issues become major challenges for data center operators. AI factories are pushing for maximum token throughput which requires the rest of the system to run efficiently, ensuring energy is focused on computation rather than cooling. Today’s liquid cooling systems are often not fully optimized from chip through to chiller. Improving how the technology cooling system loop, from cold plates to Coolant Distribution Units (CDUs), is designed can deliver significant efficiency gains. Furthermore, integrating CDUs with Building Management Systems (BMS) can also help achieve additional energy and operational savings.

The gap is twofold: (1) IT and infrastructure teams are often siloed or struggle to align on fully optimized system requirements and (2) deep expertise in liquid cooling system design and optimization remains limited.

To address these challenges, we collaborate closely with customers from initial component design through system-level implementation to ensure cross-team alignment and fully optimized solutions.

CoolIT cold plates, manifolds and CDUs are engineered for maximum performance within the smallest footprint.

Our CDUs feature advanced communication and monitoring capabilities, including support for Redfish, an increasingly common standard among major data center operators.

While high-performance hardware is essential, seamless integration into the broader system is a key focus at CoolIT. That is where we also work with operators on system design.

Mukul Girotra, Ecolab:  The shift to AI workloads represents a fundamental inflection point for data center thermal management, creating challenges that traditional air cooling simply cannot address at scale.

The reality is AI chips are generating heat densities of 40-100kW per rack, which is often 3-5x higher than traditional enterprise workloads. This creates localized hot spots that overwhelm conventional facility cooling systems and can lead to catastrophic thermal failures if not properly managed.

Here’s how we’re addressing that cooling challenge at Ecolab.

We’ve developed site-to-chip cooling management solutions that include programs for cooling water, adiabatic and direct-to-chip systems, with coolant health insights delivered via digital control platforms like Ecolab® Water Quality IQ™. 

Specifically, our 3D TRASAR™ Technology for Direct-to-Chip Liquid Cooling provides real-time monitoring of system parameters including temperature, pH, flow rates, and glycol concentration, which act as leading indicators of coolant health. These insights help operators anticipate and address degradation risk before it impacts performance and uptime.

This kind of visibility is key to managing AI workloads at scale.

Ultimately, our integrated cooling programs are designed to optimize PUE (Power Usage Effectiveness) and WUE (Water Usage Effectiveness) - helping data centers run cooler, cleaner, and more sustainably, aspects that continue to be some of the industry’s biggest challenges.  
  

Becky Wacker, Trane:  As AI workloads increasingly dominate new data center builds, operators face significant challenges in managing thermal loads and water resources. These challenges include significantly higher heat density, large, aggregated load spikes, uneven distribution of cooling needs, and substantial water requirements if using traditional evaporative cooling methods. The most critical risks include overheating, inefficient cooling systems, and water scarcity.

These issues can lead to reduced hardware lifespan, hardware throttling, sudden shutdowns, failure to meet PUE targets, higher operational costs, and limitations on where AI data centers can be built due to water constraints.

At Trane, we are evolving our solutions to meet these challenges through advanced cooling technologies such as liquid cooling and immersion cooling, which offer higher efficiency and lower thermal resistance compared to traditional air-cooling methods.

Flexibility and scalability are central to our design philosophy. We believe a total system solution is crucial, integrating components such as CDUs, Fan Walls, CRAHs, and Chillers to anticipate demand and respond effectively.

In addition, we are developing smart monitoring and control systems that leverage AI to predict and manage thermal loads in real-time, ensuring optimal performance and preventing overheating through Building Management Systems and integration with DCIM platforms.

Our water management solutions are also being enhanced to recycle and reuse water, minimizing consumption and addressing scarcity concerns.