Heat Transfer Fluids for Direct-to-Chip and Immersion Cooling Systems: How Are They Different?

As data centers, high-performance computing (HPC) systems, and advanced electronics continue to grow in power and complexity, managing heat dissipation efficiently has become one of the biggest challenges. Traditional air cooling systems, though widely used, are increasingly being replaced by more efficient solutions such as direct-to-chip and immersion cooling systems. These cooling techniques utilize specialized heat transfer fluids (HTFs) to manage heat removal. However, each cooling method requires different types of fluids that possess unique characteristics tailored to their specific thermal management needs.

1. Direct-to-Chip Cooling Systems

Direct-to-chip cooling, often referred to as liquid cooling, is an advanced technique in which the heat transfer fluid is applied directly to the surface of a chip, typically through a cold plate or heat sink that is attached to the chip. The fluid absorbs the heat generated by the chip and carries it away to a secondary loop or heat exchanger.

Heat Transfer Fluids in Direct-to-Chip Cooling

In direct-to-chip systems, the heat transfer fluid must meet specific criteria to maximize heat dissipation while ensuring the safety and longevity of the components:

• Thermal Conductivity: The fluid must have high thermal conductivity to quickly absorb heat from the chip.
• Viscosity: Low viscosity fluids are preferred to allow for efficient flow through the cooling loop, minimizing energy losses due to friction.
• Non-conductivity: The fluid should be electrically non-conductive to prevent damage to sensitive components in the event of a leak or spill.
• Compatibility: The fluid must be chemically compatible with the materials of the system, including the cold plates, piping, and seals, to prevent degradation or corrosion over time.

Common fluids used in direct-to-chip cooling include:

• Water-based coolants: These are the most common HTFs for liquid cooling, often mixed with antifreeze agents (like ethylene glycol) and corrosion inhibitors to improve performance in colder environments.
• Fluorocarbon-based fluids: These are electrically non-conductive fluids, which offer a combination of good thermal performance and safety in electronic environments.
• Glycol-water mixtures: These are commonly used for their ability to withstand low temperatures while also preventing freezing.

The effectiveness of direct-to-chip cooling relies heavily on the design of the cold plate and the thermal interface material (TIM) that sits between the chip and the plate. The heat transfer fluid helps to carry the absorbed heat away from the chip and toward the cooling loop or heat exchanger.

2. Immersion Cooling Systems

Immersion cooling systems take a different approach by completely submerging electronic components, including chips, circuit boards, or entire servers, in a heat transfer fluid. In this setup, the heat is directly transferred from the components into the fluid, which is then circulated or pumped through a system to dissipate the heat away.

Heat Transfer Fluids in Immersion Cooling

For immersion cooling, the heat transfer fluid needs to be engineered to support the specific demands of full submersion. Key attributes of immersion cooling fluids include:

• High Heat Capacity: Immersion cooling fluids require a high heat capacity to absorb large amounts of heat from components without significantly increasing in temperature.
• Non-conductivity: As with direct-to-chip cooling, immersion fluids must be electrically non-conductive to ensure that submerged electronics remain safe in case of fluid leaks.
• Stability at High Temperatures: Since immersion cooling often involves higher temperatures (compared to direct-to-chip), the fluid must be stable over time and not break down when exposed to high thermal loads.
• Low Viscosity: To ensure efficient heat transfer and fluid movement throughout the system, immersion cooling fluids should have low viscosity.

Common immersion cooling fluids include:

• Mineral oils: These are relatively inexpensive and stable oils that have been used in some early immersion cooling systems. However, they tend to have lower thermal conductivity compared to other specialized fluids.
• Synthetic oils: Fluids like, a synthetic dielectric fluid, are popular for immersion cooling because they are non-conductive, have good thermal properties, and are stable under heat.
• Fluorocarbon-based fluids: As in direct-to-chip cooling, fluorocarbons are used in immersion cooling for their non-conductive nature and high performance in terms of heat transfer.

In immersion cooling, the fluid does not need to be pumped directly to the chip, but it must circulate throughout the bath and be efficiently cooled at a separate heat exchanger. This allows immersion cooling to be particularly well-suited for high-density systems where large amounts of heat need to be managed over an extended surface area.

Key Differences Between Direct-to-Chip and Immersion Cooling Fluids

While both direct-to-chip and immersion cooling systems rely on specialized heat transfer fluids, the demands and characteristics of the fluids vary greatly due to the fundamental differences in the cooling techniques.

Application

• Direct-to-chip: Fluids are used in a loop that directly contacts the chip through a cold plate. Cooling is targeted to individual components, typically in racks or modular setups.
• Immersion cooling: Fluids are used to completely submerge the entire system or large server components. The cooling effect is spread across a large area, often within a tank or large container.

Thermal Demands

• Direct-to-chip: While these systems still require high thermal conductivity, the cooling is localized, focusing on individual chips or components. The fluid needs to quickly absorb heat and efficiently transport it away to the rest of the system.
• Immersion cooling: The fluid must handle the heat from multiple components submerged in the bath. The heat transfer demands can be higher as it involves a larger surface area and often higher heat loads across the system.

Fluid Characteristics

• Direct-to-chip: These fluids are designed to work within a loop with precise flow rates and typically have to accommodate lower thermal loads.
• Immersion cooling: These fluids are designed for full submersion and need to handle higher capacities for heat absorption while ensuring stability in a submerged environment.

System Design and Efficiency

• Direct-to-chip: In this approach, the fluid works through a closed-loop system with individual components like cold plates, which may require more complex plumbing and setup.
• Immersion cooling: The design often involves large tanks or baths, which can offer more efficient heat dissipation due to direct contact with the fluid but may require more maintenance and considerations regarding fluid purity and stability.

Conclusion

Both direct-to-chip and immersion cooling systems utilize specialized heat transfer fluids to manage the thermal challenges posed by modern, high-performance electronic components. The differences between these systems lie not only in their cooling approach—direct contact versus full submersion—but also in the properties and types of fluids they require. While direct-to-chip systems typically use fluids with high thermal conductivity and low viscosity, immersion cooling systems require fluids that can manage large amounts of heat in a more passive, bulk cooling environment. The choice of HTF depends on the cooling technique, the specific system requirements, and the desired efficiency of thermal management.

Please contact GlueRu to provide solutions to your challenges!