In-Rack Coolant Distribution Unit (CDU): The Heart of High-Density Thermal Management

The central nervous system of your high-density rack, providing precision-controlled, two-phase liquid cooling for the AI era.

As AI workloads push server power consumption beyond the limits of traditional air cooling, the infrastructure that delivers coolant to the chip becomes as critical as the chip itself. Our in rack CDU (Coolant Distribution Unit) is specifically engineered to manage the complex requirements of DI water two-phase cooling (W2PC). By decoupling the secondary IT loop from the primary facility water, our CDU ensures that your high-performance GPUs receive the exact thermal regulation they need to maintain peak performance without thermal throttling.

Key Features & Strategic Advantages

• Precision Two-Phase Management: Optimized to handle the specific flow and pressure requirements of phase-change cooling, ensuring uniform DI water delivery to every server node.

• PFAS-Free Safety: Designed exclusively for use with high-resistivity DI water, eliminating the environmental and health risks associated with chemical forever chemicals (PFAS).

• Compact In-Rack Footprint: Available in standard 4U to 8U configurations, our in rack CDU saves valuable "white space" by eliminating the need for bulky, room-scale cooling units.

• Chiller-less Compatibility: Capable of operating with high facility return water temperatures, significantly lowering your data center’s PUE and overall energy costs.

• Smart Monitoring & Leak Detection: Integrated with advanced IoT sensors for real-time flow, pressure, and resistivity monitoring, featuring automated shut-off protocols for maximum safety.

Detailed Product Introduction: The W2PC Series CDU

Our in rack CDU is more than just a pump—it is an intelligent thermal interface. In a traditional liquid cooling setup, water simply carries heat away by getting warmer. Our CDU supports the Water-to-Phase Cooling (W2PC) process, where heat is removed through the latent heat of vaporization using high-purity DI water.

By sitting directly within the server rack, the CDU minimizes the "travel distance" of the coolant, reducing pumping power and improving the response time to sudden heat spikes typical of AI training and inference. Its modular design allows for "N+1" redundancy, ensuring that even in the event of a component failure, your mission-critical AI clusters continue to operate. Whether you are retrofitting an existing data center or building a greenfield AI factory, our CDU provides the scalability needed to grow from a single rack to a massive high-density deployment.

Product Series Comparison

Targeted Application Fields

• AI Data Centers & HPC: Precise thermal management for dense GPU clusters (H100/B200/GB200).

• Edge Computing Nodes: Delivering enterprise-grade cooling in compact, remote locations.

• Enterprise R&D Labs: Enabling high-performance simulation and modeling in standard office environments.

• Modular Data Centers (MDC): The core cooling component for rapid-deployment containerized sites.

Proven Success Stories

• Hyperscale Cloud Provider: Implemented our in rack CDU to support 80kW AI racks using a DI water loop, successfully achieving a PUE of 1.08.

• SEMICON Taiwan 2025: Showcased the precision and reliability of our two-phase delivery system under high-stress exhibition conditions.

• National Research Center: Optimized a liquid-cooled supercomputing cluster using our VC-CDU series, reducing maintenance frequency by 20% while ensuring ionic purity.

💡 Still evaluating the best cooling path for your needs?

If you are planning a new data center but have not yet decided between air cooling, liquid cooling, or other more advanced architectures, please explore our dedicated service: "Cooling & Thermal System Design to Maintain." Our team, led by experts like Michael Wei, will tailor the optimal thermal management solution based on your specific facility, hardware requirements, and budget.

Consult with Our Technical Team Today