University of Notre Dame researchers will lead a new National Science Foundation- and Semiconductor Research Corp.-funded effort to develop a new area of computing titled “Extremely Energy Efficient Collective Electronics” (EXCEL).
The NSF-SRC has awarded the researchers a $4.5 million grant to fund the project over three years, which is expected to uncover fundamentally new ways of harnessing coupled dynamical systems for solving computationally hard problems in an energy-efficient way. With innovations in novel materials and devices, chip-scale system implementation, and architectural innovations and critical benchmarking, EXCEL will lay the foundation for a new computing paradigm to achieve a 1,000-fold improvement in computational energy efficiency.
“With billions of devices connected to the cloud, we have officially entered the age of data deluge,” said Suman Datta, Notre Dame Chang Family Chair of Electrical Engineering and principal investigator for the research project. “It’s imperative on us to develop sophisticated and advanced software and hardware solutions to extract key insights and actionable intelligence from all forms of data, both structured and unstructured. The primary focus of EXCEL is to develop special-purpose hardware to accelerate such data analytics in an extremely energy efficient manner. With new discoveries of emergent phenomena in solid-state materials, demonstration of new device concepts, new computational algorithms, innovative integrated circuit design techniques and new architectures, we are well-poised to lay the foundation for a radically different approach to information processing.”
The center looks to leverage brain-inspired, unsupervised learning systems to enable a highly energy-efficient, scalable computing platform. In this vertically integrated proposal, the researchers will address the theory of collective computing to rigorously establish the information capacity and computation complexity of dynamical systems, pursue physical hardware demonstration and quantify their efficacy in solving computationally hard problems that are finding ever-expanding applications in high-performance data centers, real-time cyber-physical systems and computational medicine.
The grant’s multidisciplinary team will include researchers from Georgia Tech, Pennsylvania State University, the University of California at Irvine, the University of California at San Diego and the University of Chicago.
The research project is also structured to benefit from strong engagement from industry researchers, which will facilitate technology transfer in the future. Additionally, the project also includes outreach activities that are prioritized around educating future generations of engineering students to adapt to the forthcoming evolution and revolution in information processing systems.
EXCEL will build on the research being carried out by the Center for Low Energy Systems Technology (LEAST) to form an influential driver in the quest for the next generation of energy-efficient and high-performance computing devices, circuits and systems.
Contact: Suman Datta, 574-631-8835, sdatta@nd.edu
Originally published by at news.nd.edu on October 20, 2016.