The DCX H1000 immersion enclosure is a high-density cooling solution engineered for scalable AI and HPC deployments. Designed around the industry-standard IT rack footprint (800×1200 mm), it enables seamless integration into existing data center environments without requiring fundamental infrastructure changes. The system supports standard 19”, 21”, and ORV3 server architectures with depths of up to 900 mm, ensuring broad platform compatibility. Its modular, vanity-free design allows flexible deployment configurations, including both wide and high-density narrow layouts, optimizing space utilization across different facility types, from enterprise data centers to edge and containerized installations. With a cooling capacity of up to 140 kW per rack, the H1000 delivers consistent thermal performance for next-generation high-power workloads. The enclosure architecture is optimized for efficient heat transfer and stable fluid dynamics, supporting predictable operation under sustained high loads.
DCX H1000 Immersion Enclosure
Standard Immersion Enclosure
Capacity
Up to 80 kW per server at 27°C, scaling to 140 kW per rack configuration.
Performance
Stable high-frequency operation enabled by targeted flow heat transfer without thermal throttling.
Sustainability
Reduced carbon footprint through minimized reliance on air cooling and improved thermal efficiency.
Energy Efficiency
Lower energy consumption via precise heat removal and reduced cooling overhead.
Overview
Features
- High-density immersion cooling system engineered for AI and HPC workloads in standard IT rack footprint
- Industry-standard immersion enclosure compatible with 19”, 21” and ORV3 server platforms
- Rack-based immersion system designed around 800×1200 mm footprint for seamless data center integration
- Up to 80 kW heat transfer per server and up to 140 kW per full rack configuration
- Flexible rack architecture supporting 22RU immersion + 2RU air or 44RU immersion + 4RU air setups
- Targeted flow cooling enabling higher CPU/GPU performance without thermal throttling
- Precision cooling approach reducing energy consumption and operational costs
- Integrated industrial pump system with N+1 or N+2 redundancy for high availability deployments
- Compact and modular design enabling deployment in enterprise, edge and containerized environments
- Easy access to pumps and heat exchanger compartment for simplified maintenance
Features
- Lower carbon footprint enabled by liquid cooling architecture and minimized use of HVAC systems.
- Improved server reliability through precise temperature control and elimination of hotspots.
- Extended hardware lifespan with stable thermal conditions and reduced failure rates.
- Quiet and efficient operation with minimal noise compared to conventional air-cooled systems.
- Scalable architecture supporting single, dual, and triple-level configurations for growing compute demand.
- Up to 140kW per rack in full rack configuration with exceptional heat transfer efficiency.
- Maximized ROI through higher power density and compute scaling within the same data center footprint.
- Reduced operational costs by eliminating reliance on traditional air cooling and lowering energy consumption.
System Build
Built on a compact, industry-aligned footprint (L 630 mm × W 1310 mm × H 1050 mm), the DCX H1000 is designed as a compatible immersion tank, engineered to deliver efficient, scalable cooling for high-density workloads. The enclosure integrates a clearly defined internal layout, consisting of a compute compartment for immersion-cooled servers, a dedicated networking switch compartment, an isolated hydraulic compartment, and a control panel section. This separation enables optimized thermal and mechanical management while simplifying maintenance and operational access. The hydraulic system incorporates three industrial-grade pumps configured in N+1 or N+2 redundancy, ensuring high availability and stable performance even in remote or mission-critical deployments.
Designed for flexible density configurations, the H1000 supports both 22 RU immersion-cooled + 2 RU air-cooled setups and extended 44 RU immersion-cooled + 4 RU air-cooled configurations. This allows operators to adapt the enclosure to different workload profiles while maintaining consistent cooling performance and infrastructure compatibility. The immersion-based architecture eliminates the need for high-speed airflow, resulting in significantly lower acoustic output than traditional air-cooled systems. This enables quieter operation and improves deployment flexibility in environments where noise and mechanical simplicity are critical considerations. Built with a focus on modularity and integration, the H1000 provides a robust and scalable platform for high-density computing, combining compact form factor, redundant hydraulic design, and a structured internal layout to meet the operational demands of modern AI and HPC infrastructure.
Up to 80 kW per server at 27°C, scaling to 140 kW per rack configuration.
Get in Touch
We will advise, guide and support – the ideas are always free of charge!
Get your custom solution:
We need basic information to design custom liquid cooling system:
-
Quantity and general specification and dimensions of servers, ICT systems or other to be cooled components, along with power allocation per device or rack.
-
Brief overview of your project along with your main objective: better performance, water savings, energy efficiency, heat reuse, higher density?
-
Preferred liquid cooling technology: direct chip – coldplate system or immersion solution? Planned heat rejection or heat recovery strategy.
-
Country & city nearby planed location for climate data, heat transfer performance & shipping cost estimate.