Ecolab makes AI cooling a managed service

Ecolab is positioning Cooling-as-a-Service as a managed operating model for AI data centre liquid cooling. The signal is that direct-to-chip cooling may increasingly be procured as a service tied to coolant chemistry, telemetry and field engineering rather than as standalone hardware.

• The model bundles coolant chemistry, telemetry and field engineering
• It shifts liquid cooling from hardware buying to service accountability

The fact

At Datacloud Global Congress 2026, Capacity TV interviewed Jason Walker, senior marketing manager of global data centres at Ecolab, about Cooling-as-a-Service for AI infrastructure. Walker said higher AI rack densities are pushing air cooling beyond its limits. He described an opex model that bundles cooling hardware, specialised fluids, real-time telemetry, lab checks and field engineering for direct-to-chip liquid cooling.

The Assessment

Ecolab is using the AI cooling shift to move beyond facility water treatment and into managing heat closer to the chip. The signal is not just that data centres need liquid cooling, but that many operators may lack the chemistry, cleaning and monitoring expertise to run it reliably. Ecolab is betting that hyperscalers and colocation providers will pay for predictable startup, stable operation and accountable cooling outcomes.

What to Watch

Watch for Ecolab's first large-scale hyperscale deployment metrics, especially downtime reduction, coolant stability and water-loss performance. Also monitor whether cooling hardware vendors certify their equipment for service-led models.