Researchers at the University of Minnesota will receive $3.1 million from the Department of Defense (DoD) to explore new materials, architectures, and systems that can overcome Moore’s Law, and push electronics technology forward, capable of meeting engineering, economic, and defense challenges.
Distinguished McKnight Professor and Robert F. Hartmann Chair in Electrical Engineering Jian-Ping Wang is the principal investigator of the University team and the funding is an outcome of the Electronics Resurgence Initiative (ERI) rolled out by Defense Advanced Research Projects Agency (DARPA), a DoD agency. The focus of the team will be the development of advanced magnetic tunnel junctions (MTJ) for novel computing. While the University will lead the effort, it will also be collaborating with National Institute of Standards and Technology, GlobalFoundries, and University of Arizona. Other participating University of Minnesota scientists are Prof. Tony Low and Prof. Bin Ma, both from the Department of Electrical and Computer Engineering.
DARPA’s ERI seeks to bring together research and technical expertise from academia and industry to explore new materials and architectures to bridge the gap between memory and logic functions and create new ways of computing information more quickly and efficiently. The Initiative is a $1.5 billion five-year investment designed to energize innovation in the electronics industry, and respond to anticipated challenges in the microelectronics technology industry.
One of the thrust areas under ERI is Materials & Integration which is organized into two programs: Three Dimensional Monolithic System-on-a-Chip (3DSoC), and Foundations Required for Novel Compute (FRANC). Current electronic system performance is governed by the constraints of time and power required to access memory, often referred to as the memory bottleneck. Researchers working under the FRANC program will address the bottleneck by exploring new circuit designs using novel materials and integration schemes that can minimize the movement of data between memory components and processors.
MTJs, created by sandwiching a thin insulator between two ferromagnets, have been critical to the field of spintronics; the MTJ is the basis of MRAM (magnetoresistive random-access memory), a type of non-volatile memory. The University has led research into spintronics, pursuing MTJ-based computation for more than 15 years. “In fact,” Prof. Wang notes, “our team proposed the very early idea to use magnetic tunnel junctions for the computation in random access memory (CRAM).”
With DARPA’s ERI funding, Prof. Wang and his team will work on demonstrating the value of MTJs in bridging the separation between the memory and logic functions that exist in conventional circuit designs. Their goal, under the FRANC program, is to leverage and integrate novel materials, architectures, and chip components to enhance computational speed, and performance, while lowering power consumption.