AbleCool (RDHx)-Rear-mounted chiller door

---

Ablecool

A rack-level liquid cooling solution engineered for high-density servers. It simplifies deployment, reduces energy consumption, and ensures data center stability and efficiency under AI/HPC workloads.

Key Highlights

The Ablecool replaces traditional high-power cooling methods, focusing on energy efficiency, thermal capacity, and maintainability.

Energy Saving
Up to 30% power reduction*, lowering cooling infrastructure energy use.

Thermal Capacity
Supports up to 65 kW per rack*, ideal for AI and HPC high-density deployments.

Target PUE
Enables PUE close to 1.06*, maximizing overall energy efficiency.

Modular Design
Installs on existing racks with no internal liquid loop required inside servers, simplifying deployment and maintenance.

Benefits

Efficient Thermal Management
Rack-level heat exchange minimizes hotspots and reduces dependence on traditional CRAC/chiller systems.

Flexible Scalability
Modular, per-rack upgrade path well-suited for AI clusters, HPC environments, cloud infrastructure, and 5G applications.

Reduced OPEX
Lower cooling-related power consumption and maintenance costs, while improving rack density and white-space utilization flexibility.

Core Specifications

• Mounting Type: Rear-mounted chiller door; compatible with standard 42U/48U/52U racks
• Cooling Medium: Liquid loop connected to facility CDU or chilled water system
• Max Heat Load: 75 kW (to be updated per official specification)
• Fans: Integrated variable-speed, redundant fans
• Monitoring & Control: Temperature, flow, and pressure sensors with BMS/DCIM integration
• Maintenance: Front/rear detachable access for easy cleaning and service

Applications:

AI Training & Inference
GPU-dense servers and heterogeneous acceleration platforms

High-Performance Computing (HPC)
Energy-efficient cooling for large-scale compute clusters

Cloud & Data Lake
Higher rack density with stable thermal performance

Edge Data Centers
High cooling efficiency in space- and power-constrained sites