Flexible and rapid deployment of modular liquid and airside cooling

Data center developers are in a seemingly impossible position, forced to rapidly install cooling infrastructure without always knowing tenant requirements. This article highlights this issue, positioning adaptable cooling as the only way to futureproof evolving data center needs.

Data centers are facing an existential issue: demand is outweighing supply. Deployment is expected within six to nine months, despite oftentimes not knowing how data halls will be loaded. From hyperscalers to colocation, there is a desperate need for adaptable cooling equipment that can match this need for speed.

The differences between a conventional air-cooled deployment and a liquid-cooled AI build changes everything. For developers, the challenge arises from ensuring a site can absorb changing demand without forcing expensive redesigns as the market evolves.

Traditionally, data center build programmes are long, and the technical assumptions underpinning a hall can be locked in years before the first rack even arrives. Despite this, demand pressure is pushing developers towards phased expansion and early commitments while the tenant mix is still fluid.

In the UK, CBRE’s mid-year outlook for London noted that almost all new space expected in 2025 was already accounted for before delivery, with vacancy forecasted to fall to under eight percent by year-end.

With low vacancy pushing operators and developers towards phased expansion and pre-commitment models, the market finds itself swept up in an uneasy dynamic that expects capacity to be delivered ahead of full certainty. This guessing game suddenly gets very costly, with an ideal cooling design for one tenant profile potentially becoming a constraint for another.

Divergent designs

In a January 2026 outlook, Dell’Oro Group forecast noted that thermal design powers for leading-edge GPUs are projected to exceed 4,000W by 2029, reinforcing why a growing share of AI infrastructure is being designed around liquid cooling from the outset. Alongside this, the market for cloud computing continues to rapidly grow.

Nvidia’s recently unveiled Rubin platform focuses attention on warm water, single phase direct liquid cooling, including a 45°C (113°F) supply temperature target. This change increases the window of climates where free cooling can cover a substantial portion of the year.

Ultimately, adaptable cooling is quickly proving itself to be the most futureproof and viable approach to data centre cooling system planning. A design that assumes a single operating regime can force costly compromise later, particularly in phased builds or mixed-tenant campuses.

As such, supporting air and liquid-cooled architectures within the same site strategy, with a practical means of switching between them, is fast becoming the only viable competitive option.

Switching to hybrid cooling

The predicament data centre developers are facing heavily leans towards packaged, perimeter-deployed systems as a solution. They deliver speed, being able to be installed quickly, while being highly adaptable with adjustable configurations as demand becomes clearer. This is the approach behind the Excool Switch, a packaged hybrid cooling unit designed to support both airside and liquid cooling within a single module.

The Switch is designed to serve a data hall’s need for conventional air-cooled deployments, liquid-cooled AI loads (including those with supply temperature of 45°C (113°F) and above), or a blend of both, with the presumption that the balance may change over time.

The unit combines liquid and airside cooling within a packaged module of up to 1.2MW, making 2.4MW achievable in as little as 10m of external wall space with variable load allocation between the two. It also integrates indirect glycol free cooling, supporting higher-efficiency operation when ambient conditions allow.

To facilitate rapid deployment, the units are heat-load tested prior to delivery to reduce onsite commissioning time, specifically designed for a streamlined installation process that’s supported by Excool’s short delivery times of 16-18 weeks.

This also enables late-stage deployment within the build programme. Developers can defer a significant portion of cooling capex until tenant requirements are confirmed and fit-out is imminent, improving cash flow on phased builds while still being ready regardless of whether the tenant requires liquid or air-cooled facilities.

Rapid deployment is facilitated by removing the need for primary chilled-water pipework and minimising supporting equipment requirements, giving it the lowest connected load of any comparable system. It also avoids the need to allocate internal data hall space for air cooling units or CDUs, supported by an in-built buffer tank for additional redundancy in instances of a power failure or large transient load steps.

In an operating data center, the service running on a server is often less important to facilities teams than the behaviour it creates in power and heat. A new hall is a blank canvas for a reason: it has infinite potential to accommodate anything, but cooling choices are difficult to reverse once concrete and pipework are in place.

Adaptable cooling empowers developers and operators to reduce commitment risk, while staying responsive to market demand. In a sector where the next tenant requirement may directly oppose the last, the ability to switch without rebuilding is integral to what makes a data centre commercially resilient.

In essence, the Switch delivers data centers the versatility to cater to ever-shifting tenant demands in a highly compact and cost-effective way.

For more information on the Switch unit, get in touch at [email protected]