01 / The driverWhy air runs out
Air is a poor medium for moving heat. Cooling a rack with it works to around 20 to 30 kW, but an AI rack can draw four to five times that, and you cannot push enough cold air through a cabinet to keep the chips in specification.
Water carries roughly 3,500 times the heat of the same volume of air, so the answer is to bring liquid to where the heat is. The only real question is how close to the chip you take it.
The physics
Heat removal scales with a fluid’s density and heat capacity. Liquid beats air by orders of magnitude, so a thin loop of water removes what a large volume of air cannot.
The driver: AI density
8-GPU servers at about 10 kW and GB200-class racks past 130 kW put heat on the chip that air cannot keep up with. Liquid is now a requirement, not an upgrade.
The bonus: efficiency
Liquid lets you run warm, cooling with water at 30 to 45°C, which cuts or removes mechanical chilling and pushes PUE towards ~1.1.
The trade: complexity
Water or dielectric fluid in the white space means loops, CDUs, leak management and a new operating discipline. The density and efficiency are worth it, if it is designed properly.
02 / The methodsHow close to the chip
Liquid cooling is a spectrum: from a coil on the back of an air-cooled rack, to cold plates on the chips, to submerging the whole server in fluid. Each step takes liquid closer to the heat, handling more density at better efficiency, in exchange for more change.
Best for
Watch for
Figures are indicative and vary by deployment, kit, coolant and temperatures.
The pattern is consistent: the closer the liquid gets to the silicon, the higher the density and the better the efficiency, but the bigger the change to how you build and operate. Most AI halls today land on direct-to-chip, often hybrid with air for the residual load; immersion is chosen where density and efficiency justify a wholesale change in operations.
03 / The loopThe facility loop behind the rack
Whatever the method, the heat has to get from the chip to the outside world. It does that through, usually, two loops kept apart by a CDU, a coolant distribution unit, so the clean, controlled fluid near the servers never mixes with the facility’s water.
The CDU keeps the two loops separate: clean, filtered, temperature-controlled fluid to the servers, and facility water to reject the heat.
- 01
The CDU
The heart of the system: a heat exchanger with pumps, filtration and controls that isolates the clean technology loop from facility water, regulates flow and temperature, and is sized N+1 for resilience.
- 02
Two loops, kept apart
Facility water, which can be hard, variable and shared, never touches the servers. The secondary loop is filtered, treated and tightly controlled, protecting cold plates and quick-disconnects from fouling and corrosion.
- 03
Warm water by design
Liquid cooling runs warm, often 30 to 45°C supply. Warm water rejects heat to ambient for most of the year with little or no mechanical chilling, which is where the efficiency comes from.
- 04
Quick-disconnects and manifolds
Dripless quick-disconnects and in-rack manifolds let servers be added, removed and serviced without draining the loop: the plumbing equivalent of hot-swap.
04 / The upsideEfficiency and heat reuse
Density is why you need liquid; efficiency is why you might want it before you are forced to. Because liquid removes heat at a high temperature, it changes the energy and water equation of the whole hall, and turns waste heat into something useful.
Lower PUE
Warm-water cooling cuts or removes compressor-based chilling and the fans air cooling depends on, moving PUE from about 1.5 towards about 1.1: a direct cut in energy overhead.
Free cooling, more of the year
The warmer the supply water you can use, the more hours a dry or adiabatic cooler handles the load with no chiller running. In many climates that is nearly year-round.
Water use (WUE)
Liquid cooling can reduce evaporative water use compared with some air-cooled chiller designs, but it depends on the heat-rejection choice. WUE has to be designed for, not assumed.
Heat reuse
High-grade captured heat can warm offices, district heating networks or industrial processes, turning a cost into a contribution.
The sustainability case is real and increasingly expected. A hall that runs at low PUE on warm water, ideally reusing its heat, is far easier to justify than one chilling air at 1.5 or more. See the sustainable data centre.
05 / Bringing it inRetrofitting and designing for liquid
Few operators move a whole hall to liquid overnight. The usual path is hybrid, liquid for the dense AI rows and air for the rest, and increasingly every new hall is designed liquid-ready even if it starts air-cooled. Either way, these are the things that decide whether it goes smoothly.
Leak management sits at the top: water near live electronics is the concern that stops projects, and the answer is engineering, with dripless disconnects, leak detection, drip trays, negative-pressure loops and tested procedures. Done properly, liquid cooling is no riskier than the power chain beside it. The other half is designing it with the white space and the loop together, not bolting it on afterwards.
06 / How Optronix helpsDesigned with the white space, not bolted on
Liquid cooling lives where we work: the white space, the containment, the pathways and the loop. We design and deliver liquid-ready and liquid-cooled halls, integrate CDUs and pipework with the rack layout and cabling, and retrofit density into existing air-cooled estates, with our own engineers, across Europe, North America, the Middle East and APAC.
Loop, CDUs and layout designed together
We model the loop, the CDUs, the pipework and the rack layout together, so floor loading, pathways, leak management and serviceability are designed in from the start rather than fought for later.
Matched to your density and roadmap
Rear-door retrofits, direct-to-chip rows alongside air, or liquid-ready new build, matched to your density today and your roadmap, with the residual air load planned in.
- 01
Assess and design
Density targets, method selection, loop and CDU sizing, floor loading and pathways, modelled and costed with the white space.
- 02
Install and integrate
CDUs, manifolds, pipework and containment, integrated with power and the cabling plant by our own teams.
- 03
Commission and prove
Leak-tested, flow and temperature verified, and commissioned through to integrated systems testing before any live load.
- 04
Operate and support
Operational support, smart hands and maintenance for the loop and the hall around it.
Related services
Liquid cooling
Tell us your target densities.
Retrofit or new build, we will come back with a costed liquid-cooling design: method, loop and layout, built to perform and to scale.
Sources and notes
- ASHRAE TC 9.9: liquid-cooling guidelines and facility-water temperature classes (W17 to W45).
- Accelerator and AI-system vendor specifications for rack densities and direct-to-chip heat capture.
- Common industry practice for CDUs, primary and secondary loops, immersion (single and two-phase) and heat reuse; figures are indicative and vary by deployment.
General guidance, not a design specification. Density ceilings, PUE, heat-capture and temperature figures are indicative and depend on the kit, coolant, temperatures and site; two-phase technologies and dielectric fluids are evolving and face changing environmental and regulatory scrutiny. Confirm against vendor data and an engineered design.