From cloud computing to drug discovery, we constantly require more powerful chips that produce highly concentrated heat. Extracting this heat is a major bottleneck for the next generations of computing whilst representing 30% of electricity consumption in data centers. On a global scale, this exceeds the electricity consumption of New York, and London combined. There’s an urgent need for more powerful and efficient cooling to make the future of computing more sustainable.
The current, inefficient solutions require heat to spread through many layers of packaging and interfaces before it’s extracted, resulting in poor heat transfer. At Corintis, we tackle the problem right at the core: We have developed a new precision microfluidic cooling solution, where cooling liquid flows through a network of microscopically small channels embedded in the chip. This approach can extract 10 times more heat compared to the current leading market alternatives with over 50 times higher energy efficiency. We offer a full-stack solution for chip designers that lets them tailor the design of these thousands of small cooling channels to the chip layout to ensure the right amount of cooling gets delivered to the right location.
The team at Corintis and EPFL have experimentally demonstrated the massive performance and efficiency improvement that this technology can offer, which resulted in traction from some of the world’s biggest chip designers and data center operators. In this project, we will tackle the next step in market entry, by qualifying the reliability of our technology for long-term commercial use. This project enables us to conduct an in-depth reliability study on our microchannel cooling technology to address the long-term risks associated with clogging and thermal cycling. This is currently a key factor that is preventing us from advancing our technical discussions with customers to the next step.
Persons involved in the project
Last update to this project presentation 30.01.2023