Cooling Module
Heat-source contact, internal flow path, sealing, tubing and defined connections.
MODULE SUPPLIERServer Liquid Cooling Guide: From Modules to Rack Interfaces
Server liquid cooling moves processor heat into a coolant loop. A closed-loop module rejects that heat through a radiator to server airflow; an externally connected system uses rack or facility cooling equipment. This guide explains the design inputs and tests needed for each arrangement.
View Server Module Scope
From Cooling Module to Facility: Which Layers Apply?
A closed-loop module works within the server and rejects heat to air. Externally connected liquid cooling adds rack distribution and facility interfaces. The four layers below help identify the equipment and integration work your project needs.
Server Node
Board and chassis fit, airflow, service, controls and real compute workload.
SERVER OEM / INTEGRATORRack Distribution
Manifolds, node balancing, isolation, monitoring and service procedures.
EXTERNAL LIQUID SYSTEMSFacility Interface
CDU, facility water, heat rejection, site controls and operating policy.
SITE COOLING TEAMChoose the Cooling Path from the Workload
Rated power is only a starting point. Selection should use sustained and peak workloads, concurrent heat sources, allowed temperatures, server density, available airflow, service requirements and the infrastructure the site can support.
| Path | Useful when | Questions to resolve |
|---|---|---|
| Air cooling | Heat and acoustic targets remain manageable in the available chassis airflow. | Fan power, inlet temperature, altitude, filters and component hotspots. |
| Closed-loop liquid-to-air module | Processor heat can be moved to an internal radiator while the server continues to reject heat to air. | Radiator space, chassis airflow, pump and fan control, and residual component cooling. |
| Externally connected direct-to-chip | The server can connect to suitable rack or external liquid-cooling equipment. | Cold plates, coolant, flow distribution, connections, CDU or other heat-rejection equipment. |
Chassis, Flow and Heat-Rejection Interfaces
Integration succeeds when mechanical, hydraulic and service inputs are reviewed together. A module cannot be assessed independently from its installation and external operating conditions.
- Controlled board and processor revision, socket or PCB, keep-out zones and mounting loads
- Chassis height, adjacent components, residual airflow and installation sequence
- Tube routes, bend space, connections, leak-impact zones and service access
- Coolant and material compatibility, inlet conditions, flow and pressure-drop budget
- Pump, radiator, CDU or facility boundary, controls and abnormal-operation ownership
- Transport orientation and the maintenance procedure used in the field
Plan Leakage, Condensation and Layered Validation
Specify leak-test conditions and check condensation risk against coolant temperature, ambient humidity and dew point. Test the module first, then verify the installed server and any external cooling system.
Fit, sealing and function
Check controlled dimensions, assembly, defined leak/pressure condition, flow function and agreed thermal evidence.
Workload and integration
Run the real node configuration with airflow, control, service and abnormal-state conditions.
Distribution and operations
Validate balancing, CDU or facility interfaces, monitoring, isolation and operating procedures.
Build a Comparable Server Cooling RFQ
Give each supplier the same platform files, operating conditions, quantity and test requirements. Ask what is included, what you must provide and which details still need confirmation.
- Server form factor, board, processor and configuration revisions
- Sustained, peak and concurrent heat-source workloads
- Mechanical, fluid, electrical and service boundaries
- Test questions, conditions, acceptance rules and required records
- Prototype, pilot and production quantities, timing and packaging
- Change-control and revalidation expectations
| Supplier review | Evidence to request |
|---|---|
| Engineering | Input-gap list and boundary drawing |
| Manufacturing | Controlled process and inspection plan |
| Validation | Condition, method, result and disposition |
| Change control | Approved baseline and impact review |
Server Cooling Questions Before Project Review
For model examples and customization options, see the Server Liquid Cooling Modules page.
Does ThermalBro provide complete rack or facility cooling infrastructure?
ThermalBro’s documented core scope is server liquid cooling modules and related engineering and manufacturing. Rack distribution, CDU and facility-water equipment must be defined with the customer and the responsible infrastructure suppliers.
What information is needed for a server module review?
Provide the server form factor, board and processor platform, heat sources and workloads, chassis files, routing and connection constraints, external-loop conditions, validation goals, quantities and target project stage.
How should thermal test conditions be defined?
State the workload, duration, ambient and inlet conditions, coolant, flow or pressure condition, fan and pump control, temperature measurement points, sample configuration and acceptance rule.
What is the difference between module-level and rack-level validation?
Module validation checks agreed fit, sealing, flow and thermal evidence for the supplied module. Rack-level validation also includes node interaction, distribution, CDU or facility interfaces, controls and operating procedures owned by the system and site teams.
Method sources: Open Compute Project Cold Plate Cooling and OCP cold-plate loop/qualification publications. External references explain methods; they do not certify ThermalBro products.
Discuss Your Server Cooling Requirements
Share the board and chassis files, heat sources, workload and expected quantity. Tell us whether the server uses an internal radiator or connects to external cooling equipment.