Who makes liquid-cooled servers?

When I first visited a hyperscale data center using liquid cooling, I could hear an unusual silence. Gone was the roar of fans—the servers were cooled by quiet, efficient liquid loops. This experience changed how I saw modern server design.

Liquid-cooled servers are manufactured by companies such as Dell Technologies, HPE, Lenovo, Supermicro, Inspur, and specialized cooling partners like Vertiv, Schneider Electric, CoolIT Systems, and LiquidStack.

They are designed to handle high-power workloads—especially AI and HPC—where traditional air cooling is no longer enough.

What is a liquid-cooled server?

At its core, a liquid-cooled server is not very different from a regular one. What sets it apart is how it manages heat.

A liquid-cooled server uses a circulating coolant—usually water or dielectric fluid—to absorb heat from CPUs, GPUs, or memory modules through cold plates or immersion baths, then transfers that heat to a radiator or heat exchanger for dissipation.

How it works

1. Heat Absorption:
   A cold plate attached to a processor transfers heat to the coolant through direct contact.

2. Fluid Circulation:
   A pump moves the warm fluid through a closed loop, delivering it to a heat exchanger or cooling distribution unit (CDU).

3. Heat Rejection:
   The fluid releases heat through radiators or facility water lines, then returns cooled to repeat the cycle.

Cooling Method	Description	Typical Use
Direct-to-Chip (D2C)	Coolant flows through cold plates mounted on CPUs/GPUs	Data centers, HPC
Immersion Cooling	Servers submerged in dielectric fluid	AI training clusters
Rear-Door Heat Exchanger	Air-cooled servers with liquid-cooled doors	Retrofits and hybrid centers

Modern data centers use hybrid approaches—combining liquid and air—to handle variable workloads while reducing energy usage.

What benefits does it offer to data centers?

In my experience working with data center engineers, the shift to liquid cooling is driven by both necessity and efficiency. Air simply cannot keep up with the heat loads of today’s AI systems.

Liquid-cooled servers deliver superior thermal efficiency, reduced energy use, higher density per rack, quieter operation, and improved sustainability for data centers.

1. Higher cooling efficiency

Liquid can absorb heat up to 3,000 times faster than air, which means it can cool higher TDP processors without overheating. This is critical for AI, rendering, or machine learning workloads.

2. Higher rack density

Traditional air-cooled racks handle about 5–10 kW. Liquid-cooled racks easily support 30–80 kW or more. This allows more computing power in the same space.

Feature	Air Cooling	Liquid Cooling
Max Rack Density	10 kW	80 kW+
Cooling Efficiency	Moderate	High
Energy Cost	Higher	Lower
Noise Level	Loud	Low
Space Efficiency	Low	High

3. Lower energy consumption

Because liquid removes heat more efficiently, chillers and fans consume less power. Many data centers report up to 20–30% total energy savings after switching.

4. Better reliability and performance

Stable operating temperatures reduce CPU/GPU throttling, improving uptime and extending hardware lifespan.

5. Sustainability

Liquid cooling enables heat reuse—the waste heat can warm nearby buildings or feed district heating networks. It also reduces water use compared to large-scale air chillers.

6. Quieter and smaller footprint

Fewer high-speed fans mean quieter operations, and compact liquid loops make more efficient use of data hall space.

How to choose a server cooling provider?

Selecting a reliable liquid cooling partner is crucial. Over the years, I’ve seen data centers succeed or fail based on whether they matched the right technology to their facility.

To choose a server cooling provider, evaluate your thermal load, verify vendor experience, confirm compatibility with your infrastructure, and review energy savings, service plans, and safety features.

Step 1: Assess your workload

Determine whether your servers run CPU-intensive, GPU-heavy, or mixed loads. This affects whether you need direct-to-chip cooling, immersion cooling, or a hybrid system.

Step 2: Check technology compatibility

Ask providers how their systems integrate with your racks, PDUs, and existing HVAC. Compatibility ensures easy retrofitting and minimal downtime.

Step 3: Evaluate vendor credibility

Leading manufacturers include:

• Dell Technologies – PowerEdge XE9640 with direct-to-chip liquid cooling
  
• Lenovo Neptune™ – A long-running liquid cooling platform
  
• HPE – Offers direct liquid cooling with Cray and Apollo systems
  
• Supermicro – Customizable servers supporting liquid-ready nodes
  
• Inspur & Huawei – Large-scale AI and hyperscale liquid-cooled racks
  
• Vertiv, Schneider Electric, CoolIT Systems, LiquidStack – Infrastructure providers supplying CDUs, manifolds, and immersion tanks
  

Vendor Type	Examples	Focus
Server OEMs	Dell, HPE, Lenovo, Supermicro	Integrated liquid-ready systems
Cooling Specialists	Vertiv, Schneider Electric, CoolIT Systems	Facility infrastructure
Immersion Innovators	LiquidStack, Submer, GRC	Full immersion and edge setups

Step 4: Review efficiency and cost

Request energy models showing potential PUE (Power Usage Effectiveness) improvement. A well-designed liquid system can cut your PUE by 0.05–0.1 points.

Step 5: Verify safety and maintenance

Ensure your provider offers:

• Leak detection systems
  
• Flow and pressure sensors
  
• Maintenance contracts for coolant replacement
  
• Operator training and ²⁴⁄₇ service
  

Step 6: Consider scalability and roadmap

Choose a vendor whose roadmap supports future processors and increased rack density. You don’t want to redesign your system every time a new chip generation launches.

What are the trends in enterprise cooling tech?

Liquid cooling is changing faster than ever. The rise of AI, HPC, and sustainability initiatives is driving a wave of innovation in how data centers manage thermal energy.

Current trends in enterprise cooling focus on direct-to-chip architectures, full immersion systems, AI-based control, modular CDUs, eco-friendly coolants, and hybrid liquid-air integration.

1. Direct-to-chip cooling

This approach routes coolant directly to processors using microchannel cold plates. It’s now used in most AI-optimized servers, allowing up to 80 kW per rack.

2. Immersion cooling

Servers are fully submerged in dielectric fluids. This technique removes nearly all air-based cooling needs, improving energy efficiency and reducing maintenance.

3. Hybrid air-liquid cooling

Many facilities are transitioning gradually. They combine rear-door heat exchangers with liquid loops, creating a hybrid system that cools high-density nodes while retaining existing air systems.

4. AI-driven thermal control

Artificial intelligence is now managing data center temperatures in real time. Sensors feed data to AI algorithms that optimize flow rate, fan speed, and pump pressure dynamically.

5. Modular and scalable CDUs

Compact coolant distribution units now handle hot-swappable connections and smart monitoring. Modular CDUs make it easier to expand capacity rack by rack.

6. Eco-friendly and long-life coolants

Next-generation fluids are non-toxic, biodegradable, and have low electrical conductivity. They last longer and reduce environmental impact while improving safety.

Trend	Description	Key Advantage
Direct-to-Chip Cooling	Liquid loops target CPUs and GPUs directly	High density efficiency
Immersion Cooling	Servers submerged in dielectric fluids	Simplified thermal control
AI Optimization	Smart pump and fan regulation	Real-time efficiency gains
Modular CDUs	Scalable liquid infrastructure	Easy expansion
Eco Coolants	Sustainable, non-toxic fluids	Environmental compliance

7. Waste heat recovery

Liquid cooling allows heat recapture at higher outlet temperatures (40–60°C). Some European data centers now use this energy to heat offices or residential zones.

8. Edge and micro data centers

Compact edge data centers use liquid cooling to fit powerful compute clusters in small footprints. These systems reduce latency for AI inference or 5G applications.

9. Standardization and collaboration

Industry groups like the Open Compute Project (OCP) are establishing open standards for liquid interfaces and manifolds, making cross-vendor deployment easier.

10. The future: autonomous cooling ecosystems

The next generation of enterprise cooling will feature self-regulating liquid systems. They’ll automatically balance flow, refill coolant, detect leaks, and adjust performance through AI monitoring—all with minimal human input.

Conclusion

Liquid-cooled servers are transforming how data centers handle modern workloads. Built by leading brands like Dell, Lenovo, HPE, and Supermicro—alongside infrastructure experts like Vertiv and Schneider—they deliver high density, energy savings, and sustainability. Choosing the right provider means evaluating compatibility, safety, and scalability. With AI-driven control, modular CDUs, and eco-friendly fluids leading the way, liquid cooling is fast becoming the new standard for enterprise computing.