Back to News
IndustryAI Understanding briefing

Ecolab unveils liquid‑cooling ecosystem to accelerate AI data‑center builds

Ecolab announced a new open AI‑infrastructure ecosystem that pairs its CoolIT liquid‑cooling technology with Siemens power solutions, Alfa Laval heat exchangers and Dassault Systèmes virtual‑twin tools to speed AI‑data‑center design, cut water and energy use, and shorten time‑to‑token.

4 min readRead the linked source
Source-provided image accompanying Ecolab unveils liquid‑cooling ecosystem to accelerate AI data‑center builds
Source referenceSource recorded
Publisher
ecolab.com
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Key terms

Compute
The processing resources required to train and run models, often measured in FLOPS or GPU hours.
Token
A chunk of text processed by language models, such as a word piece or symbol.

What happened

Ecolab launched an open AI‑infrastructure ecosystem at the SC26 conference, bringing together Siemens, Alfa Laval and Dassault Systèmes to deliver an integrated, liquid‑cooling‑centric solution for high‑density AI workloads. The partnership combines Ecolab’s CoolIT CHx™ 3000 coolant distribution units (CDUs) with Siemens’ power‑distribution, automation and digital‑twin capabilities, while Alfa Laval supplies high‑efficiency heat‑exchangers, pumps and valves. A virtual‑twin showcase, co‑developed with Dassault Systèmes, models the full data‑center thermal and water flows to optimise “water per ,” energy efficiency and deployment speed. The ecosystem is positioned as a blueprint for data‑center operators to move faster from design to production, improve uptime, and reduce water and energy consumption.

At the SC26 conference on November 17, 2026, Ecolab announced the formation of an open AI‑infrastructure ecosystem aimed at accelerating the design‑to‑production cycle for AI data centers. The ecosystem is built around Ecolab’s CoolIT CHx™ 3000 coolant distribution units, which provide direct‑to‑chip liquid cooling for high‑density AI workloads.

Siemens will contribute power‑distribution hardware, automation software and digital‑twin capabilities, creating a unified power‑and‑cooling solution that eliminates the traditional siloed approach to data‑center design. The partnership is intended to simplify deployment, reduce thermal bottlenecks, and shorten the time required to bring a new AI model to operational status (the “time‑to‑”).

Alfa Laval joins as the heat‑exchanger and flow‑component partner, supplying high‑efficiency heat exchangers, pumps and valves that integrate with Ecolab’s thermal loop management. This collaboration focuses on improving heat‑transfer performance, reliability, and the ability to reuse waste heat within the data‑center ecosystem.

Dassault Systèmes provides a virtual‑twin platform that links technology cooling system (TCS) and facility water system (FWS) models. The twin simulates the entire cooling chain—from rack design inputs through CDU selection—to optimise water use, coolant performance and power efficiency. Metrics such as water per , token efficiency and time‑to‑token capacity are used to evaluate design trade‑offs.

Ecolab framed the ecosystem as a blueprint for data‑center operators, promising faster deployment, higher uptime, and measurable sustainability benefits. The company highlighted its broader corporate goals of protecting water resources and reducing environmental impact, positioning the ecosystem as part of its 2030 sustainability targets.

Source details: ecolab.com

Why it matters

The announcement targets a critical bottleneck in AI : the rapid scaling of high‑density data‑center infrastructure. By integrating liquid cooling directly with power and automation systems, Ecolab and its partners aim to cut deployment complexity and accelerate “time‑to‑,” a metric that reflects how quickly a new AI model can be trained and generate value. The ecosystem also promises measurable sustainability gains—lower water use per token and reduced energy draw—addressing growing regulatory and corporate pressure to make AI compute greener. If adopted, the approach could set a new industry standard for AI‑focused data‑center design, influencing hardware vendors, cloud providers and enterprises that need to spin up large AI clusters quickly and responsibly.

AI model training now consumes a growing share of global electricity and water resources. Traditional air‑cooled data‑center designs struggle to keep pace with the thermal density of modern GPUs and ASICs, leading to higher energy use and longer deployment cycles. By integrating liquid cooling directly with power and automation, the ecosystem could materially lower the energy‑per‑ and water‑per‑token ratios, addressing both cost and environmental concerns.

The “time‑to‑” metric is increasingly important for enterprises that view AI as a competitive advantage. Faster deployment translates to quicker model iteration, earlier revenue generation, and reduced capital expenditure. An ecosystem that shortens this timeline could become a differentiator for cloud providers and large enterprises seeking to stay ahead in AI capabilities.

Sustainability pressures are mounting from regulators and investors. Demonstrating concrete reductions in water and energy consumption can help data‑center operators meet ESG targets and avoid potential penalties. The ecosystem’s focus on circular water use and heat‑re‑use aligns with broader industry moves toward greener .

If the integrated solution proves scalable, it may shift market dynamics away from legacy air‑cooled vendors toward liquid‑cooling specialists and system integrators that can offer end‑to‑end designs. This could spur further innovation in coolant chemistry, heat‑exchanger efficiency, and digital‑twin‑driven design workflows.

The involvement of major players—Siemens, Alfa Laval and Dassault Systèmes—adds credibility and may accelerate adoption across the supply chain, from hardware manufacturers to cloud service providers.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Interactive Concept Check+10 Points
AI Models Explained Quiz

Which component of an AI application is the machine-learning model itself?

What to watch next

Key indicators will include early‑adopter pilots that publish performance data (e.g., water‑per‑, deployment timelines), the rollout of the CoolIT CHx™ 3000 CDUs in commercial data‑centers, and any follow‑on agreements between Ecolab and major cloud operators. Watch for regulatory commentary on water‑use metrics for AI and for competitor responses from traditional air‑cooled cooling vendors. The evolution of the virtual‑twin tool’s adoption by design firms will also signal how quickly the integrated workflow gains traction.

Pilot deployments: Look for announcements from data‑center operators or cloud providers that adopt the CoolIT CHx™ 3000 CDUs and publish performance metrics such as reduced water‑per‑ or faster deployment timelines.

Commercial availability: Track Ecolab’s product rollout schedule, pricing details (currently undisclosed), and any announced service‑level agreements for the ecosystem components.

Regulatory response: Monitor water‑use reporting requirements for AI in key jurisdictions (e.g., EU, US states) that could affect adoption incentives.

Competitive moves: Watch for responses from traditional air‑cooled cooling vendors or other liquid‑cooling firms that may launch alternative integrated solutions.

Virtual‑twin uptake: Assess how quickly design firms and data‑center engineers adopt the Dassault Systèmes twin platform, which could become a standard design tool if it demonstrably improves efficiency.

Related guides & quizzes

Found this useful?