Daikin Applied said on 10 November that it has introduced Reference Architecture D60-W, a 60-megawatt cooling blueprint for AI factories based on the NVIDIA DSX AI Factory Platform with NVIDIA Vera Rubin NVL72. The company says liquid cooling handles 92% of the load and air cooling the remaining 8%.
By our arithmetic, that split puts 55.2 MW of the 60 MW IT load on liquid and 4.8 MW on air, and the six deployable units average 10 MW each. Daikin Applied says its CDU and data center dry cooler are under development, with availability targeted for the second half of 2027, about eight months from now by our count.
Business Pill · Vertical Integration
A short explainer on vertical integration. It teaches the general idea and does not discuss Daikin Applied or D60-W.
The key insight: As we read it, Daikin Applied’s release sells cooling as one integrated plant rather than as separate machines: by our arithmetic 55.2 MW of the 60 MW design runs on liquid, and the company says dry coolers replace evaporative cooling towers, with its own CDU and dry cooler targeted for the second half of 2027.
The company says D60-W gives data center developers, engineers and operators “a starting point for designing cooling from rack to ambient”, connecting cooling equipment and controls into a single system.
According to the release, the design runs two temperature loops: a high-temperature liquid loop serves the coolant distribution units (CDUs), and a dedicated low-temperature loop serves the air-cooling rack containment units. The rack containment units are DDC Solutions S-Series units, the release says.
D60-W is built around Daikin Applied’s oil-free Magnitude WME magnetic bearing centrifugal chillers, with capacities up to 10 MW, the company says. It says oil-free technology removes oil heaters, pumps and standard oil servicing requirements.
IT load by cooling method in Daikin Applied’s D60-W reference architecture: liquid cooling handles 92% of the 60 MW load and air cooling 8%, per the company. Megawatt figures (55.2 MW and 4.8 MW) and the 10 MW per-unit average across six deployable units are our arithmetic.
Water and Operating Modes
Daikin Applied says D60-W switches between dry cooling, hybrid cooling and full mechanical cooling based on outdoor conditions, and that in dry cooling mode the chiller compressors are off. The release says dry coolers carry the cooling load for most of the year in suitable climates.
Heat rejection relies on dry coolers instead of evaporative cooling towers, which the company says “eliminates cooling-tower make-up water”. Our earlier coverage tracked AI data centers above 250 MW and rack power plans for 2027.
“AI factories are designed to turn energy into output as efficiently as possible, and cooling infrastructure sits right at the center of that challenge”, said Yu Nishiwaki, president and CEO of Daikin Applied. Karim Shalash, director of Data Center Solutions, said: “An AI factory is a system of systems”.
The Structural Read
As we read the release, the design is sized to a named NVIDIA platform: Daikin Applied says D60-W is based on the NVIDIA DSX AI Factory Platform with NVIDIA Vera Rubin NVL72, and it scales by deployable unit, six of them for 60 MW.
As we read the release, two parts of the system are not yet available: Daikin Applied says its CDU and data center dry cooler are under development, while its Magnitude WME chillers, with capacities up to 10 MW, anchor the design.
Karim Shalash, director of Data Center Solutions, Daikin Applied, 10 November 2026
“A cooling decision at the rack carries implications for water temperatures, flow requirements, plant operation and heat rejection”
Three Implications
FOR DATA CENTER DEVELOPERS Daikin Applied says D60-W gives engineering teams a framework covering rack, plant, heat rejection and controls, with six deployable units of 10 MW each on average by our arithmetic.
FOR WATER USE Daikin Applied says heat rejection uses dry coolers instead of evaporative cooling towers, which eliminates cooling-tower make-up water.
FOR TIMING Daikin Applied says its CDU and data center dry cooler are targeted for the second half of 2027.
The framework describes its bottom layer, Energy & Physical, as: “Gigawatts, transmission, cooling, land, water, rare-earth magnets.”
As we read it, D60-W sits in that layer: Daikin Applied says the design handles a 60 MW IT load and replaces cooling-tower make-up water with dry coolers. The framework predates the D60-W release and does not discuss it.
What Is Not Established
The release gives no price for D60-W, names no customer or site using it, and gives no water usage effectiveness figure. It does not say how many hours a year the dry-cooling mode runs in any named climate.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Daikin Applied says its D60-W blueprint cools a 60 MW Vera Rubin AI factory with liquid handling 92% of the load, 55.2 MW by our arithmetic, and dry coolers in place of cooling towers, with its own CDU and dry cooler targeted for the second half of 2027.
A note on sourcing. We read Daikin Applied’s press release of 10 November 2026 on PR Newswire in full. The megawatt split, the per-unit average and the months to availability are our arithmetic from the figures it gives. We did not contact Daikin Applied, NVIDIA or DDC Solutions.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
