Microchannels cool high performance processors

1 min read

As part of the CarriCool project, Fraunhofer researchers have worked with IBM to develop what is said to be an effective method for cooling processors. According to the team, by integrating microchannels into the silicon interposer, a significant increase in performance can be achieved. Meanwhile, the team has also integrated passive components for voltage regulators, photonic ICs and optical waveguides into the interposer.

The team, from Fraunhofer’s Institute for Reliability and Microintegration (IZM) in Berlin and Dresden developed an interposer with microchannel structures and hermetically sealed vias. The structure allows coolant to be pumped through the microchannels, channelling heat away from the processor.

“Until now, the cooling structures have not been very close to the computer core itself, which means the coolers are mostly applied from above,” says Dr Hermann Oppermann, group leader at Fraunhofer IZM. “The closer you get to the heat source, the better the temperature can be limited or the output increased. In high performance computing in particular, data rates are increasing continuously. Therefore, it is important to have effective cooling in order to ensure a higher clock rate. Previous cooling systems were not so effective in this context. With this new cooling system, performance can be increased significantly.”

The challenge was not only to integrate the small channels into the interposer, but also to seal them hermetically. The solution is an interposer featuring horizontal cooling channels and vertical channels for the electrical connections. In order to prevent contact between the water and the electrical vias, each individual contact is specially sealed.

The researchers also integrated voltage regulators for the power supply, as well as optoelectronic components for data transmission into the interposer. “By combining interposer, cooling, voltage regulators and optical interconnection technology, we have reached a new level of integration that allows smaller circuits with more power,” said Dr Oppermann. “This is an important step in high-performance computing, as we achieve higher clock speeds in the same amount of space.”