Hongming Sheng manufactures rack-level liquid cooling components for AI servers and data center systems, including copper cold plates, thermal modules, machined cooling structures and related metal assemblies.

We support custom components for GPU, CPU and power-module cooling according to customer drawings, rack layouts and mechanical requirements.

Copper Cold Plates for AI Servers

Copper cold plates are used to transfer heat from GPUs, CPUs and power modules into a liquid cooling loop.

For server applications, both thermal structure and mechanical fit need to be considered during manufacturing.

• Contact surface flatness
• Cooling channel geometry
• Mounting hole positions
• Inlet and outlet locations
• Sealing surfaces
• Overall assembly dimensions

GPU & CPU Thermal Modules

Cooling modules may combine cold plates, heat sinks, seals, brackets, frames and connection structures into one assembly.

• GPU cold plate assemblies
• CPU cooling components
• Power module thermal parts
• Mounting frames and brackets
• Cooling support structures
• Custom machined interfaces

CNC Cooling Channels & Sealing Surfaces

Cooling channel geometry, sealing surfaces and mounting interfaces are important features in rack-level liquid cooling components.

• CNC-machined flow channels
• Controlled sealing surfaces
• Threaded ports and mounting features
• Flat contact areas
• Interface dimensions for adjacent components

Rack-Level Cooling Component Support

Rack-level cooling systems can require multiple machined and fabricated metal parts working together.

• Cold plates
• Cooling module frames
• Mounting brackets
• Thermal support structures
• Metal housings
• Precision machined interfaces

Manufacturing Capabilities

Key Rack Integration Points

For rack-level liquid cooling projects, the cooling component must fit both the heat source and the surrounding server structure.

• GPU / CPU mounting interface
• Cold plate envelope dimensions
• Port and connector orientation
• Clearance with server chassis parts
• Bracket and fastener positions
• Assembly and maintenance access
• Rack-space limitations

Typical Applications

• AI data center cooling systems
• High-density server racks
• GPU computing platforms
• CPU liquid cooling systems
• Power module cooling
• High-performance computing equipment
• Power electronics thermal systems

Quality & Assembly Inspection

• Material verification
• Critical dimensional inspection
• Mounting hole position checks
• Cooling channel and feature inspection
• Flatness checks when specified
• Sealing surface inspection
• Assembly fit verification when required
• Leak or pressure testing when specified by the project

What to Send for Quotation

• 2D drawing or 3D CAD model
• Material requirement
• Required quantity
• GPU / CPU interface information
• Critical dimensions and tolerances
• Flatness requirement if applicable
• Channel and port information
• Assembly or rack interface requirements
• Leak or pressure test requirements if applicable

Confidential OEM Manufacturing

Customer drawings, cooling layouts, server structures and technical information are treated as confidential. NDA support is available for AI server and data center liquid cooling projects.