Director, Computational Systems & Technology Division
Research Associate Professor, Ming Hsieh Department of Electrical Engineering
Ph.D. Electrical Engineering, USC
M.S. & B.S. Computer and Electrical Engineering, Purdue University
crago at isi.edu
Dr. Stephen Crago is the Deputy Director of Computational Systems and Technology (CS&T) at the Information Sciences Institute, where he has been since 1997, and holds a joint appointment as a Research Associate Professor in the Ming Hsieh Department of Electrical Engineering. As Deputy Director of CS&T, Dr. Crago has primary responsibility for CS&T’s Arlington contingent in Arlington, Virginia, and has managerial responsibilities in reconfigurable computing and wireless in addition to his own group’s research in high-performance cloud and multi-core software. Since joining ISI, Dr. Crago’s research has been sponsored by DARPA, NASA, ONR and other government agencies, and Dr. Crago has led many projects, from small, focused research efforts to large multi-disciplinary collaborations, across the range of his research interests. He received his B.S. in Computer and Electrical Engineering and his M.S. in Electrical Engineering from Purdue University and his Ph.D. in Computer Engineering from the University of Southern California.
Dr. Crago’s current research interests include heterogeneous computing, high-performance and embedded cloud computing, introspective systems, and parallel software. As power density and device physics limit our ability to scale homogeneous multi-core processors, heterogeneity and efficiency must increase if we are to continue to get more processing capability on anything from embedded platforms to clouds for big data processing. For applications and software to adapt to these changing platforms, we will need to develop software that can target heterogeneity and run-time systems that can adapt to dynamic workloads while handling the complexity of highly parallel, heterogeneous systems. Dr. Crago’s research group is focused on the exploration and development of software and techniques to exploit highly parallel, heterogeneous systems efficiently, in terms of performance, power, and productivity.