Cooling architecture Technology

Hybrid air and liquid cooling

Liquid takes the heat off the processors while air handles the rest of the rack. Almost every liquid-cooled hall works this way.

hybridair coolingliquid coolingtransition

Direct answer

What this record says

Hybrid air and liquid cooling is a scaling cooling architecture using liquid at the processors, cooled air for everything else. It is typically applied at Broadly applicable from 30–150+ kW racks. Last reviewed Aug. 9, 2026 against 1 source.

1 sourceReviewed Aug. 9, 2026
Maturity
Scaling
Coolant
Liquid at the processors, cooled air for everything else
Typical density
Broadly applicable from 30–150+ kW racks
Suppliers tracked
4 suppliers
MaturityScaling
Cooling mediumLiquid at the processors, cooled air for everything else
Typical applicationBroadly applicable from 30–150+ kW racks

Architecture

How heat moves

Hybrid cooling runs two heat paths at once. Cold plates remove heat from graphics processing units and central processing units into a rack liquid loop, while fans carry heat from memory, drives, power supplies, and network devices into room air. The liquid passes through a coolant distribution unit (CDU) to facility water; the air passes through a rear-door unit or room air handler to a coil.

It fits phased retrofits, colocation halls, and mixed fleets. Its main risk is treating residual air heat as a small fixed number. Liquid capture changes by server model and workload, while neighboring legacy racks still shape airflow.

The facility needs both water and air capacity, coordinated controls, leak and dew-point management, and fault plans for each path. Ask the server maker for heat-to-liquid and heat-to-air data at expected workloads, including switches and power shelves. Model first-day and full-build conditions, then test what happens if a CDU, room fan, chilled-water branch, or sensor fails.

Procurement should define control ownership across information technology and facilities, meter both paths, and preserve enough air turndown to avoid wasted fan energy. Compare whole-site power and usable rack capacity, not liquid capture alone.

01Two paths at once: chip to liquid, and other components to air

Why teams choose it

  • Lets a site move from air to liquid in stages instead of all at once
  • Keeps working with a mixed fleet of old and new hardware
  • Puts liquid only where it earns its cost

What to validate

  • Two cooling systems have to be sized, controlled, and maintained together
  • How much heat the liquid captures changes with each server generation
  • The heat still going into the room can be larger than teams expect

Connected records

Companies in this category

View all companies in this category

Connected records

Required equipment

View all required equipment

Connected records

Tracked deployments

View all tracked deployments

Evidence ledger

Sources and evidence

The links below show where the factual claims came from. Supplier specifications remain supplier-reported unless the record names independent operating evidence.

  1. 01
Record information
Record ID
DCC / TECH / HYBRID-COOLI
Record reviewed
Aug. 9, 2026
Record first published
Aug. 9, 2026
External sources
1
Search visibility
Offered to search engines

Latest coverage

Intelligence mentioning Hybrid air and liquid cooling

View all intelligence mentioning hybrid air and liquid cooling

Decision support

Guides and comparisons mentioning Hybrid air and liquid cooling

Direct answers

Frequently asked questions

Direct answers drawn from the record, its comparison fields, and the evidence linked below.

What is Hybrid air and liquid cooling?

Hybrid cooling runs two heat paths at once. Cold plates remove heat from graphics processing units and central processing units into a rack liquid loop, while fans carry heat from memory, drives, power supplies, and network devices into room air. The liquid passes through a coolant distribution unit (CDU) to facility water; the air passes through a rear-door unit or room air handler to a coil. It fits phased retrofits, colocation halls, and mixed fleets. Its main risk is treating residual air heat as a small fixed number. Liquid capture changes by server model and workload, while neighboring legacy racks still shape airflow. The facility needs both water and air capacity, coordinated controls, leak and dew-point management, and fault plans for each path. Ask the server maker for heat-to-liquid and heat-to-air data at expected workloads, including switches and power shelves. Model first-day and full-build conditions, then test what happens if a CDU, room fan, chilled-water branch, or sensor fails. Procurement should define control ownership across information technology and facilities, meter both paths, and preserve enough air turndown to avoid wasted fan energy. Compare whole-site power and usable rack capacity, not liquid capture alone.

Where is Hybrid air and liquid cooling typically used?

Broadly applicable from 30–150+ kW racks. We class the approach itself as scaling.

What are the main advantages of Hybrid air and liquid cooling?

Lets a site move from air to liquid in stages instead of all at once. Keeps working with a mixed fleet of old and new hardware. Puts liquid only where it earns its cost.