Products & Tools

Schneider Electric pitches software layer to run liquid cooling across AI data centres

Schneider Electric has proposed a supervisory software layer called liquid cooling orchestration to manage chillers, pumps and cold plates across AI data centres as rack densities and water budgets tighten.

Schneider Electric backs liquid cooling software for AI
Schneider Electric backs liquid cooling software for AITho-Ge / pixabay
By Elena Vasquez5 min read

Why it matters

  • Schneider Electric outlined its liquid cooling orchestration approach on 19 November 2026.
  • The supervisory software would coordinate chillers, pumps, heat rejection and airflow across AI facilities, sitting above local controls.
  • The software reads NVIDIA GPU MAX-LPS, MAX-P and MAX-Q profiles to size cooling capacity, and uses GPU and rack power telemetry to predict heat load.
  • Schneider Electric said running the full heat chain as one system recovers stranded cooling capacity, cuts pumping energy and increases economiser hours.
  • The company tied the approach to improving power usage effectiveness and water usage effectiveness across multi-site AI fleets.

Schneider Electric wants a supervisory software layer to run liquid cooling across entire AI data centres, managing chillers, pumps and cold plates as one coordinated system rather than a pile of separate assets. The company laid out the approach on 19 November 2026, calling it liquid cooling orchestration, and it arrives as rack power densities climb and liquid cooling shifts from exotic to standard practice in AI facilities.

The stakes are straightforward. Operators now have to match cooling systems to fast-changing AI workloads without wasting energy or water. Schneider Electric said orchestration software can help facilities maintain thermal headroom, make better use of cooling capacity and support GPU performance — three things that increasingly decide whether an AI data centre runs efficiently at all.

The company's argument reflects a wider shift in the sector. Operators are moving from conventional enterprise facilities to larger AI-focused campuses with denser racks and more complex power and thermal requirements. Cooling, once managed largely as a local facilities function, is becoming tightly linked to IT workload planning and hardware performance.

Why can't local controls handle AI cooling on their own?

The problem is physics and timing. AI factories typically rely on a hybrid thermal design: liquid-to-chip cooling for high-density AI systems, plus conventional air cooling for building infrastructure and other IT equipment. The equipment list is long — coolant distribution units, pumps, liquid distribution networks, computer room air handlers, fan walls, chillers, dry coolers, cooling towers and adiabatic heat rejection systems.

These systems do not all respond at the same speed. AI workloads can shift rapidly, but the thermal and mechanical equipment supporting liquid cooling operates over slower timescales. Schneider Electric said that mismatch creates a need for higher-level coordination rather than simple reactions to every change in chip activity.

Under the proposed model, local controls would still manage equipment such as coolant distribution units within expected operating ranges. The orchestration software sits above them and adjusts setpoints across the broader cooling chain. Schneider Electric said this would let operators coordinate chillers, heat rejection equipment, pumps and airflow systems using both historical and real-time operating data.

The aim is to ensure enough cooling is available before sustained heat loads affect the coolant loop. Predictive analytics could anticipate thermal demand, flag abnormal conditions and suggest supervisory changes that improve overall facility efficiency.

How does the software connect to GPUs?

This is the most concrete part of the proposal. Schneider Electric said the software can be tied directly to compute behaviour, and its example uses NVIDIA GPUs. Orchestration takes account of MAX-LPS, MAX-P and MAX-Q profile configurations when assessing how much cooling capacity is needed to support a required level of performance.

The monitoring loop works like this. The software tracks GPU and rack power telemetry as an early indicator of heat load, and correlates that information with manifold return temperature and flow. Predictive models then recommend — or automatically make — adjustments to coolant distribution units and chilled water plant settings before thermal demand rises.

That last point matters for operators watching both their power bills and their hardware. Cooling that arrives late forces conservative settings; cooling that arrives early lets facilities run lean.

What does managing the full heat chain buy operators?

Schneider Electric's core claim is that running the full heat transfer chain as a single system unlocks efficiency that piecemeal control leaves stranded. The company put it directly:

"Managing the full heat transfer chain as one system, from the cold plate to heat rejection, lets operators run warmer supply temperatures, gain more economizer hours, cut pumping energy, and recover stranded cooling capacity without compromising thermal resilience or GPU performance."

Each element of that sentence maps to a cost line. Warmer supply temperatures and more economiser hours reduce chiller work. Lower pumping energy trims the electricity bill. Recovered cooling capacity defers capital spending on new equipment.

The company also linked the approach to power usage effectiveness and water usage effectiveness — two measures data centre operators increasingly track as AI infrastructure expands. Better visibility into cooling requirements, it said, could help owners make more accurate decisions on facility design, construction and infrastructure investment, including estimates of likely power and water demand for cooling at different locations. According to Schneider Electric, a more detailed view of operating requirements could reduce unnecessary capital spending and lower operating costs.

What about mixed air-and-liquid sites and multi-site fleets?

A key operational headache is the growing number of mixed air and liquid environments inside AI facilities. Schneider Electric said orchestration software can ingest telemetry from both types of cooling infrastructure and present a unified operational view, allowing coordinated control across systems that may support several server platforms with different thermal profiles.

The company then scaled the argument up. As AI data centre fleets expand across regions, cooling optimisation is becoming a fleet issue rather than a single-site engineering task, Schneider Electric said. A centralised view of thermal performance, infrastructure utilisation, energy efficiency and water consumption would let operators compare sites, identify weaker assets and standardise operating practices.

That fleet framing is where the software pitch meets the market reality. AI campus buildouts are accelerating, and operators choosing control architectures now will shape their efficiency and water metrics for years.

Why this matters beyond Schneider's product line

Schneider Electric is one of the largest suppliers of data centre power and cooling infrastructure, so its push to define cooling as a software coordination problem signals where the industry expects operations to go. If liquid cooling orchestration becomes standard practice, cooling decisions move closer to workload scheduling — and further from the facilities department.

The company's position paper stops short of announcing a shipping product with a release date or price. But its direction is clear: as racks get denser and water and power budgets get tighter, the cooling layer becomes an orchestration problem, run from software, across whole fleets.

Source: itbrief.com.au

Share this article:

More from Elena Vasquez

Elena Vasquez

Show full bio

Market editor covering technology and product news in artificial intelligence and machine learning at AI In Context.

744 articles

Related articles

  1. Voters Approved 40 of 41 Ballot Measures Restricting AI Data Centers
  2. Lenovo Bets on Liquid Cooling and Services to Solve AI's Power Problem
  3. Modine Opens 682,000-Square-Foot Chiller Factory in Texas
  4. Startup Drogue raises $20M to power floating data centers with ocean heat
  5. Element Six, Entegris plan synthetic diamond for chip cooling

« Previous articleNext article »