In Rack CDU: Localized Liquid Cooling for AI & HPC Racks

All Categories
EN EN
Localized Coolant Distribution for High-Density Server Racks

Localized Coolant Distribution for High-Density Server Racks

An in rack coolant distribution unit provides localized control and circulation for liquid cooling within a high-density server rack. Positioned close to the IT equipment, the unit can manage coolant delivery between facility-side infrastructure and server-level cooling loops while reducing the distance between the cooling source and heat-generating components. Depending on system requirements, an in rack coolant distribution unit can incorporate pumps, heat exchangers, valves, sensors, controls, filtration, and distribution connections. Configuration can be matched to rack thermal load, coolant flow, supply and return temperatures, pressure requirements, fluid compatibility, and available installation space. This architecture is suitable for AI servers, GPU clusters, HPC systems, cloud infrastructure, and data centers deploying direct-to-chip or other rack-level liquid cooling technologies.
Get A Quote

Case Study

Localized Cooling for AI Servers

An AI computing deployment required localized coolant management for a rack containing multiple high-power GPU servers. An in rack coolant distribution unit was integrated between the facility cooling infrastructure and server-level liquid loops. The configuration coordinated coolant delivery to GPU cold plates through rack manifolds, hoses, and quick disconnect couplings. System planning considered rack thermal load, required flow rate, supply and return temperatures, operating pressure, coolant compatibility, and available rack space. Locating the distribution equipment close to the servers helped organize coolant connections and service access within the rack. The modular architecture also allowed additional GPU nodes to be connected as computing capacity increased without redesigning the entire cooling infrastructure.

HPC Rack Cooling

A high-performance computing rack required a controlled liquid cooling interface for multiple CPU and accelerator servers. The in rack coolant distribution unit was configured to manage coolant circulation between the facility loop and technology-side cooling assemblies. Manifolds distributed liquid to individual cold plates, while sensors could monitor flow, pressure, and temperature at key points. The design accounted for processor heat load, coolant flow requirements, pressure drop, connection layout, and maintenance access. By consolidating coolant management inside the rack, the system provided an organized connection point for several liquid-cooled servers. The configuration could also be adapted as additional HPC nodes were added to the rack or computing density changed.

Data Center Liquid Cooling Upgrade

A data center planned to introduce liquid-cooled servers into existing high-density racks while maintaining compatible facility infrastructure. An in rack coolant distribution unit provided a localized interface between the facility-side cooling supply and server liquid loops. The deployment used distribution connections, hoses, quick disconnects, and monitoring components to manage coolant delivery to processor cold plates. Design considerations included rack heat density, coolant flow capacity, supply and return temperatures, pressure limits, piping routes, and service clearance. This approach allowed operators to implement liquid cooling at selected racks rather than converting the complete facility simultaneously. Additional distribution capacity could be planned as more AI and HPC equipment was introduced.

Related products

IP65 Insertion Type Digital Glass pH Electrode RS485 Output Direct Immersion pH Sensor for CDU Cooling Water Monitoring ADP2400 VIEW DETAILS

IP65 Insertion Type Digital Glass pH Electrode RS485 Output Direct Immersion pH Sensor for CDU Cooling Water Monitoring ADP2400

FMX280 Multi-Parameter Magnetic Flow Meter with Temperature Pressure Monitoring Tri Clamp Digital Mag Flowmeter for CDU Rack VIEW DETAILS

FMX280 Multi-Parameter Magnetic Flow Meter with Temperature Pressure Monitoring Tri Clamp Digital Mag Flowmeter for CDU Rack

High Precision Stainless Steel Vortex Flow Meter DN25-DN150 4-20mA Pulse RS485 Liquid Flowmeter for Liquid Cooling CDU FVX260 VIEW DETAILS

High Precision Stainless Steel Vortex Flow Meter DN25-DN150 4-20mA Pulse RS485 Liquid Flowmeter for Liquid Cooling CDU FVX260

High Precision 0.2% Digital Pressure Transmitter with LED LCD Display 4-20mA RS485 Pressure Sensor for Liquid Cooling CDU PX300 VIEW DETAILS

High Precision 0.2% Digital Pressure Transmitter with LED LCD Display 4-20mA RS485 Pressure Sensor for Liquid Cooling CDU PX300

An in rack coolant distribution unit is designed to manage liquid cooling within or adjacent to a high-density server rack. The unit can act as a localized interface between facility cooling infrastructure and server-level cooling loops, distributing coolant to CPUs, GPUs, or other high-power components through cold plates and rack manifolds. Depending on the application, the configuration may include pumps, heat exchangers, valves, sensors, filtration, control systems, supply and return connections, and quick disconnect interfaces. Important design parameters include rack thermal load, coolant flow rate, supply and return temperatures, operating pressure, fluid compatibility, connection size, installation space, and monitoring requirements. The unit can support direct-to-chip cooling in AI servers, HPC clusters, cloud infrastructure, and enterprise data centers. Its rack-level architecture provides a structured approach to coolant management while allowing cooling capacity to be configured around specific equipment and deployment requirements.

Frequently Asked Questions

What is an in rack coolant distribution unit?

An in rack coolant distribution unit manages coolant circulation and distribution close to liquid-cooled server equipment. It can connect facility-side cooling infrastructure with rack-level or server-level cooling loops, delivering coolant through manifolds, hoses, and other connections to CPU or GPU cold plates.
An in rack CDU can regulate and distribute coolant between the facility cooling loop and technology-side equipment. Depending on the configuration, it may include pumps, heat exchangers, valves, sensors, controls, and filtration. The unit helps manage flow, temperature, pressure, and coolant delivery within a high-density rack.
Yes. An in rack coolant distribution unit can be configured for GPU servers used in AI and HPC applications. The system can distribute coolant to multiple GPU cold plates through manifolds and quick disconnects. Capacity should be selected according to rack thermal load, required flow, pressure, and temperature conditions.
Depending on system requirements, an in rack CDU may include pumps, heat exchangers, valves, sensors, filters, controllers, supply and return manifolds, and connection interfaces. Some applications require additional monitoring or fluid management components. The final configuration depends on the rack architecture and facility cooling infrastructure.
Selection should consider rack thermal load, coolant flow rate, supply and return temperature, operating pressure, coolant type, heat rejection requirements, physical dimensions, connection layout, and maintenance access. For AI or HPC deployments, future rack expansion and the number of liquid-cooled servers should also be considered during sizing.

Related article

Intelligent Power Distribution, Precision Metering – An In-Depth Look at the GLGW PDU Series

21

May

Intelligent Power Distribution, Precision Metering – An In-Depth Look at the GLGW PDU Series

View More
Connecting the World, Powering the Future

21

May

Connecting the World, Powering the Future

GLGW delivers high-quality cable solutions that keep the world connected. Built on trust and driven by innovation, our products provide reliable, efficient connectivity for industries, businesses, and communities. Experience the power of seamless connections and discover how GLGW is shaping a connected tomorrow.
View More
How to Strip Fiber Cable Without Damaging Fibers?

07

Aug

How to Strip Fiber Cable Without Damaging Fibers?

Stripping a fiber optic cable requires extreme precision. Because the glass cladding and core are extremely thin and fragile, even a microscopic scratch or nick can lead to severe signal loss (attenuation), microbends, or complete fiber failure und...
View More
Fiber Cable Tensile Strength for Aerial Installation.

24

Aug

Fiber Cable Tensile Strength for Aerial Installation.

Deploying optical fiber cables in aerial spans across long distances—such as utility poles over valleys, mountain ridges, and expansive rural infrastructure—exposes the cable assembly to continuous mechanical, gravitational, and environ...
View More

Customer Testimonials

Mark Wilson

The rack-level CDU gave our team a practical interface between facility cooling and multiple GPU servers. The organized connections simplified coolant distribution and helped with service planning.

Eric Morgan

We integrated the CDU into an HPC rack with several liquid-cooled nodes. The centralized distribution arrangement made flow monitoring and server connection management easier.

Get a Free Quote

Our representative will contact you soon.
Email
Mobile/WhatsApp
Name
Company Name
Message
0/1000
coolant distribution unit coolant distribution units coolant distribution unit for data center coolant distribution unit cdu coolant distribution liquid to liquid coolant distribution unit coolant distribution unit for HPC coolant distribution unit for AI server rack level coolant distribution unit in rack coolant distribution unit
Localized Coolant Management

Localized Coolant Management

An in rack coolant distribution unit brings liquid cooling management closer to the servers that generate heat. Instead of routing every server connection directly to distant facility equipment, the rack-level unit can provide a centralized interface for coolant supply and return. Depending on the system architecture, it can manage circulation, heat exchange, flow control, temperature monitoring, and distribution across multiple liquid-cooled nodes. Manifolds can divide coolant between CPU and GPU cold plates, while sensors monitor operating conditions at selected points. The configuration can be designed around rack thermal load, flow requirements, pressure limits, coolant properties, and available space. This localized architecture helps organize complex liquid cooling connections within high-density computing environments.
Direct-to-Chip System Integration

Direct-to-Chip System Integration

In rack coolant distribution units can support direct-to-chip cooling architectures by connecting server-level cold plates with the broader liquid cooling infrastructure. A typical arrangement may include GPU or CPU cold plates, hoses, quick disconnect couplings, rack manifolds, pumps, and the distribution unit. Depending on the facility design, the unit may also interface with a heat exchanger or facility water circuit. Key operating parameters include coolant flow rate, supply temperature, return temperature, pressure, and fluid compatibility. Centralizing these connections at rack level can make it easier to organize multiple liquid-cooled servers and establish defined service points. This architecture is particularly relevant to AI, HPC, and other high-density computing deployments.
Scalable Rack-Level Deployment

Scalable Rack-Level Deployment

As AI and HPC workloads increase, individual server racks can require significantly greater thermal management capacity. An in rack coolant distribution unit provides a modular foundation for expanding liquid cooling around these high-density environments. A rack can be configured for a selected number of liquid-cooled servers and expanded as additional nodes are introduced. System planning can evaluate current and future thermal loads, CDU capacity, coolant flow, heat rejection, connection requirements, and available rack space. Standardized supply and return connections, manifolds, hoses, and quick disconnects can help simplify integration across compatible equipment. This approach allows liquid cooling to be deployed rack by rack, supporting phased data center upgrades while maintaining a structured architecture for future computing expansion.

Get a Free Quote

Our representative will contact you soon.
Email
Mobile/WhatsApp
Name
Company Name
Message
0/1000
Ningbo GLGW Nova Materials Technology Co., Ltd.

Copyright © 2026 by Ningbo GLGW Nova Materials Technology Co., Ltd. - Privacy policy