AMCool - About the project

Global trends of electrification and digitalisation are fuelled by rapid developments in semiconductor technology and continue to change the way we work, travel, and connect. As the power density of the hardware behind these trends continues to increase, thermal management becomes the bottleneck – and AMCool aims to tackle this challenge.

The DTI-coordinated research project “Advanced Module Cooling (AMCool)” addresses the issue of rapidly growing waste heat generation in semiconductor-based technologies. Traditional cooling concepts and their associated manufacturing methods have fallen behind the requirements of modern electronics, leading to higher peak temperatures, lower efficiency, and shorter component lifetimes. In response to this challenge, AMCool aims to develop highly efficient two-phase cooling concepts, leveraging the possibilities of additive manufacturing (AM), and contribute to a more sustainable, electrified future.

In recent years, AM has emerged as a compelling solution to thermal management challenges. In contrast to traditional manufacturing techniques, such as machining, brazing, and skiving, AM excels at producing complex shapes that can adhere to a specific thermal field rather than manufacturing constraints. Furthermore, AM allows to combine multiple parts into a monolithic structure, which eliminates assembly steps, failure points and thermal barriers, while at the same time improving recyclability. Capitalizing on these inherent advantages, AMCool sets out to develop novel, AM-fabricated cooling solutions for two relevant future industries: data centres and electric vehicles.

Generating impact where it matters

The first use case of AMCool are traction inverters for electric vehicles (EVs). A traction inverter contains several power modules which transform alternating current into direct current and vice versa. This process happens both during charging and driving periods and generates significant heat losses. New generations of power electronics semiconductor materials, for instance silicon carbide (SiC) or gallium nitride (GaN), could potentially carry higher currents at superior efficiency; however, the associated heat generation surpasses the capacity of conventional cooling approaches for power modules.

The successful development and implementation of a compact, highly effective cooling solution through the AMCool project will pave the way for widespread adoption of these new semiconductor materials. On the one hand, this will enable better overall system efficiency, greater range and longer component lifetime for the individual EV. On the other hand, adequately cooled SiC and GaN power modules constitute a crucial step towards the economic viability of decentralised energy-storage approaches, such as vehicle-to-grid.

The project’s second use case concerns server cooling for high-performance computing and AI applications. As the power density per server rack has grown from <30kW to above 100kW over the span of only a few years, a new generation of cooling infrastructure is needed. A significant part (up to 40%) of a data centre’s total energy expenditure is spent on cooling alone; therefore, more efficient cooling concepts will have a noticeable easing effect on the enormous strain that data centres are putting on Denmark’s energy grid today.

While the data centre industry is still struggling with the transition from air cooling to direct liquid cooling, AMCool goes a step further and exploits the benefits of two-phase cooling for current and future generations of high-end processors. Successful implementation of two-phase technology in a data centre context will drastically lower energy demand for cooling, improve power usage efficiency, and contribute to longer component lifetime and lower failure rates. These are all necessary steps to ensure Denmark’s competitiveness in a data-driven world while maintaining compliance with sustainability targets.

Project facts

The AMCool project runs from January 2025 to December 2027 and is supported by EUDP – The Danish Energy Technology Development and Demonstration Program. The cumulative project budget is DKK 16.7 Mio. The consortium consists of four partners: Danish Technological Institute, Aalborg University, Danfoss Climate Solutions, and Heatflow ApS.