DOE funds new research to advance computer chip technology

The Department of Energy (DOE) has selected a multidisciplinary team that includes Cornell to advance a superconducting approach to advanced computer chip technology. The team will explore ways to use new superconducting materials and structures in ultra-energy-efficient Superconducting Digital (SCD) electronics aimed at emerging artificial intelligence and quantum computing technologies.

Their project, “Advanced superconducting integration process enabling sustainable hardware for AI and quantum computing,” is one of 11 multidisciplinary peer-reviewed projects selected by DOE to receive a total of $73 million in investments to accelerate new technologies from discovery to commercialization. The project is being funded through the Accelerate initiative through DOE’s Offices of Advanced Scientific Computing Research and Nuclear Physics.

The team is led by DOE’s Thomas Jefferson National Accelerator Facility and in addition to Cornell includes imec and the New York Center for Research, Economic Advancement, Technology, Engineering, and Science (NY CREATES). The team’s wide-ranging strengths include concept and design, fundamental material development, and process development and implementation, with Cornell contributing expertise in characterization of materials.

Team member Katja Nowack, assistant professor of physics in the College of Arts and Sciences, has significant experience in the characterization of superconducting materials. Her toolbox of techniques enables imaging the properties of these materials so they can be better tuned to application at scale.

“My lab brings expertise in a pretty unique way to image the relevant materials and structures,” Nowack said. “We specialize in a type of magnetic imaging at low temperatures that we typically use to explore emergent phenomena in quantum materials. But it so happens that our imaging is well-suited to help develop a greater understanding of the fundamentals of the materials involved in SCD electronics. This is a critical need to advance the kind of technology we pursue in this project.”

Having identified a key technology gap that currently hinders advancing digital superconductor electronics, the team’s approach is to innovate major changes, including scalability, to the existing fabrication process.

The team seeks to develop improved superconducting and barrier materials that would perform better and tolerate higher processing temperatures, making them fully compatible with the fabrication processes of conventional electronics technology.

In a large-scale application such as AI running in a data center, superconducting materials offer a 100X improved energy efficiency, even after accounting for the overhead of a closed-loop cryogenic system for cooling. Superconducting digital logic chips (suitably modified) can also be placed near quantum computing chips, to control and communicate with them. That means that this technology is important for enabling both quantum and classical computing in the future.

DOE’s larger goal with its Accelerate initiative is to fuel innovation in basic research so that new materials and technologies can be scaled and transitioned into new products and capabilities to enhance the country’s economic health and security.

“This research will integrate novel concepts and approaches in use-inspired basic research to address gaps or challenges that limit the ultimate transition to applied research for further development and demonstration,” said Asmeret Asefaw Berhe, director of DOE’s Office of Science, in announcing the grants. “Achieving these research goals will greatly accelerate the innovation cycle, which currently can take years to decades to realize.”

More news

View all news
		Gold surface of a computer chip