Vol.005 - Not Ready to Move to Liquid Cooling? How EVAC Sustains Air Cooling in High-Density Data Centers

 

When Conventional Air Cooling Reaches Its Limit: How EVAC Opens Up New Thermal Headroom for High-Power Platforms

In thermal design, heat is never merely a secondary concern; it is a critical factor that directly affects system performance, stability, and reliability. Within a conventional thermal architecture, TIM materials help eliminate air gaps at the interface, heat pipes transfer heat, heatsinks increase the available area for heat dissipation, and fans carry heat away from the system.

This approach has long supported the thermal requirements of servers and other high-performance equipment. However, as AI computing continues to drive demand for greater processing capability, chip power consumption has risen accordingly, and conventional air-cooling architectures are gradually approaching their practical limits.

Why Has EVAC Gained Attention?

Since 2019, platform TDP levels from Intel, AMD, and later NVIDIA have continued to increase. In conventional thermal designs, a common approach is to place heat pipes beneath the heatsink and extend them upward or outward to maximize the effective airflow exposure area, thereby improving overall cooling performance.

Yet as power levels continue to rise, this approach inevitably encounters limitations. Once power exceeds 350W, conventional air cooling offers very little remaining thermal margin. At that point, the issue is no longer simply whether the heatsink is large enough, but whether heat can be transported more effectively to locations where it can be dissipated more efficiently. This is precisely why EVAC has emerged as a new option for high-power platforms.

What Is EVAC?

EVAC stands for Extended Volume Air Cooling. As the name suggests, it extends the effective cooling space of the heatsink into other areas of the system for heat exchange.

Its purpose is not merely to enlarge a single heatsink, but to extend the capability of air cooling beyond a confined local area through a more comprehensive thermal path and airflow-path design. From a design perspective, EVAC focuses on increasing heat transfer area, optimizing airflow distribution, and improving overall heat exchange efficiency.

This design logic means that air cooling is no longer limited to the small area directly above the CPU. Instead, it incorporates the entire usable system space and available airflow conditions into the thermal design strategy.

What Kind of Problems Can EVAC Solve?

As chip TDP continues to rise, the most common challenges faced by conventional air cooling are insufficient local space and limited heat transfer area. Even when a vapor chamber is added at the base of the heatsink and heat pipes are extended upward or outward, overall cooling performance can still be constrained by mechanical limitations and airflow conditions.

The value of EVAC lies in the fact that it does not simply add more material. Rather, it redistributes thermal and airflow paths so that heat can be exchanged across a larger volume. For high-power platforms, this creates renewed thermal headroom within an air-cooling architecture that would otherwise be close to its limit.

In other words, EVAC does not solve a single-component problem. It addresses the air-cooling bottleneck of high-power platforms at the system level.

What Environments Are Particularly Suitable for EVAC?

EVAC is not limited to a specific industry. It is particularly suitable for applications where high wattage is required, yet the system must still rely on an air-cooling architecture.

As processor power continues to increase and conventional air cooling approaches its limit, EVAC becomes a more practical extension when the system must remain air-cooled for reasons such as risk control, deployment convenience, or broader system architecture considerations. Another ideal scenario for EVAC is when there is still room inside the chassis for customized mechanical design.

Because the core of EVAC lies in reconfiguring thermal and airflow paths, it extends cooling capability from a confined local region into a larger heat exchange volume. As a result, if the mechanical design still allows adjustment of airflow paths, module placement, and cooling layout, EVAC can deliver greater design value.

For applications that must balance high-power processor cooling with deployment efficiency, EVAC is not simply a way to make the heatsink larger. It is a solution that enhances thermal flexibility and overall system feasibility while remaining within an air-cooling framework. Therefore, any system environment facing high heat density, requiring an air-cooling architecture, and retaining a certain degree of mechanical integration flexibility is particularly well suited to benefit from EVAC.

EVAC Solutions

 

To address different platform and power requirements, REGO has developed EVAC solutions for both Intel and AMD platforms. From 1U to 2U form factors and from 300W to 500W TDP levels, different models can be selected according to mechanical constraints, platform compatibility, and target cooling capacity.

Platform	Model	Form Factor	TDP	Socket
Intel	RGB121-P-EX1U-1	EVAC 1U Passive Cooler	300W	LGA4677
Intel	RGB121-P-EX2U-1	EVAC 2U Passive Cooler	350W	LGA4677
Intel	RGB220-P-EX2U-1	EVAC 2U Passive Cooler	400W	LGA4677
AMD	RGE110-P-EX2U-1	EVAC 2U Passive Cooler	350W	SP5
AMD	RGE220-P-EX2U-1	EVAC 2U Passive Cooler	350W	SP5
AMD	RGE110-P-EX2U-2	EVAC 2U Passive Cooler	500W	SP5

 

 Conclusion

As AI and high-performance computing continue to drive chip power upward, the challenges facing conventional air cooling are becoming increasingly apparent. The significance of EVAC is not that it replaces conventional air-cooling logic altogether, but that it creates additional cooling headroom by making use of available system space once conventional air cooling approaches its limit.

By expanding the heat transfer area and establishing a more efficient heat exchange path based on system airflow conditions, EVAC enables high-power platforms to remain stable within an air-cooling architecture. For high-power platforms, this is not merely a change in heatsink form factor, but a thermal strategy that balances performance and risk control.