CDU 1-MW

High-Capacity Liquid Cooling. Ready When You Are.

Purpose-built for rapid deployment of high-density AI infrastructure.

Download Datasheet
Talk to an expert

Deployed in real-world AI environments and designed to support next-generation GPU platforms.

The CDU-1MW delivers reliable, high-performance liquid cooling for modern AI workloads. Designed for speed and simplicity, it enables operators to deploy liquid cooling infrastructure without delays – while integrating seamlessly into existing data center environments, including facility water systems and hybrid cooling architectures.

Key Features

1

Designed for Direct-to-Chip Liquid Cooling

Rack-based form factor, narrow approach temperatures, elevated operating water temperatures, and intelligent onboard controller.

2

High Reliability

The LiquidStack coolant distribution unit is equipped with dual, redundant pumps to ensure N+1 operation and deliver high flow and high available head pressure. The pumps can operate on a rotational schedule or by load sharing, enhancing system reliability and efficiency. It’s also USA manufactured and factory tested prior to shipping.

3

User-Friendly Design

Our coolant distribution units for liquid cooling feature dual, front-mounted electrical panels for easy access and redundant operation.

4

Easy Transport & Installation

Casters, leveling feet, forklift pockets and floor anchoring brackets and a compact IT-rack form factor ensures easy installation in both new and retrofit environments. The versatility of the LiquidStack data center CDU allows placement in-row or at the perimeter, accommodating various data center layouts.

Purpose-Built CDU Benefits for High-Density AI Workloads

Rapid deployment

Quick-ship availability accelerates time-to-live

High-density support

Designed for demanding GPU workloads

Compact footprint

Integrates easily with typical rack form factor

Proven performance

Tested and deployed in real-world AI applications

AI training clusters

Inference at scale

Retrofit deployments

Capacity expansion projects

Work with our experts to configure a CDU-1MW deployment tailored to your performance, scale, and facility requirements.

 Start your consultation

Featured Resources

Frequently Asked Questions

  • Why Direct-to-Chip Cooling for AI

    Direct-to-chip liquid cooling delivers coolant directly to CPUs and GPUs via cold plates, removing heat at the source without submerging hardware. It maintains standard server form factors, supports retrofit and hybrid environments, and offers the most practical path to high-density liquid cooling for most AI deployments. Some residual airflow is still required for non-chip components, making hybrid architectures common in practice.

    Learn More
  • Air vs Liquid Cooling for AI Infrastructure

    Air cooling is practical for low-density environments but struggles beyond approximately 30–50 kW per rack, making it unsuitable for modern AI workloads. Liquid cooling delivers significantly higher heat transfer efficiency, supports rack densities of 100 kW and above, and reduces energy consumption at scale. This article compares both approaches across density, efficiency, and deployment readiness to help you determine when the switch to liquid cooling is justified.

    Learn More
  • CDUs vs Rear Door Heat Exchangers

    CDUs and rear door heat exchangers (RDHx) both address the limits of conventional air cooling, but serve different roles. RDHx units extend existing air-cooled infrastructure with minimal server changes, making them a viable transitional solution at moderate densities. CDUs enable true direct-to-chip liquid cooling, supporting rack densities above 100 kW and providing a scalable path from rack-level to hyperscale deployments.

    Learn More