Daikin Applied Unveils 60MW AI Factory Cooling Design

The D60-W reference architecture pairs liquid and air cooling across six deployable units built on the NVIDIA DSX AI Factory Platform.

Updated on Nov. 14, 2026 in Data Centers

Isometric editorial illustration of six identical matte cooling units in a straight line on a concrete pad, linked by coolant pipes, illustrating a data center cooling design.
Daikin Applied unveiled a 60 MW cooling reference design for AI factories built on the NVIDIA DSX AI Factory Platform, pairing liquid and air cooling across six deployable units. AI Illustration. Upload story photo >

Daikin Applied announced a 60 MW cooling reference architecture for AI factories, designed around the NVIDIA DSX AI Factory Platform with NVIDIA Vera Rubin NVL72. The D60-W supports a 60 MW IT load across six deployable units, handling 92% of that load with liquid cooling and 8% with air cooling.

Why it matters

AI facilities concentrate large amounts of computing load, making heat removal a central design constraint for data center builders. A pre-engineered cooling blueprint gives operators a standardized template for high-density deployments in the United States.

The architecture runs a high-temperature liquid loop for coolant distribution units and a separate low-temperature loop for air-cooling containment units. Oil-free Magnitude WME magnetic bearing centrifugal chillers, rated up to 10 MW, anchor the design, which uses dry coolers instead of evaporative towers.

The players

Daikin Applied

A division of the Japanese HVAC manufacturer Daikin Industries that supplies commercial and industrial cooling systems, including chillers and data center thermal equipment.

The details

The D60-W coordinates rack containment units, coolant distribution units, chillers and dry coolers through a supervisory layer called iDCM, operating them as one plant. The system shifts between dry cooling, hybrid cooling and full mechanical cooling based on outdoor conditions, switching compressors off in dry mode and using waterside economisers to pre-cool return water in hybrid mode.

Timeline

  1. November 14, 2026: Daikin Applied announced the D60-W cooling blueprint.

The Tech Race

The D60-W is built to integrate with NVIDIA's DSX AI Factory Platform and Vera Rubin NVL72 racks, placing Daikin Applied alongside other infrastructure vendors competing to supply the cooling layer of NVIDIA-based AI facilities. Its reliance on dry coolers over evaporative towers reflects a design direction that reduces reliance on water-intensive heat rejection.

For operators planning AI-ready facilities, the D60-W offers a pre-engineered cooling template that could shorten design work for high-density deployments. Its hybrid and dry-cooling modes also indicate how outdoor climate conditions may shape the energy profile of such builds.

The takeaway

High-density AI computing is pushing cooling design toward mixed liquid and air systems that adapt to weather. Operators evaluating new AI facilities should weigh how a cooling architecture handles seasonal and climate variation alongside its raw capacity.

Further reading

Explore more data center infrastructure coverage for context on how AI facilities are being built and cooled.