Dynamic Thermal Estimation and Management in Nanoscale Systems

Temperature will be a limiting factor in the performance of future computer systems. This talk focuses on developing techniques for estimation and management of temperature using information estimation and control theoretic concepts. Using limited and noisy on chip thermal sensors, estimation of the chip thermal profile presents a rich vein of theoretical and practical challenges. Furthermore, autonomously controlling the thermal profile while maintaining acceptable performance is a major challenge.
Ankur Srivastava is currently an Associate Professor in the ECE department with joint appointment with the Institute for Systems Research at University of Maryland, College Park. He received his B.Tech in Electrical Engineering from Indian Institute of Technology Delhi in 1998, M.S. in ECE from Northwestern University in 2000 and PhD in Computer Science from UCLA in 2002. His primary research interest lies in the field of high performance, low power electronic systems and applications such as computer vision, data and storage centers and sensor networks. He has published numerous papers on these topics at prestigious venues. He has been a part of the technical program committee of several conferences such as ICCAD, DAC, ISPD, ICCD, GLSVLSI. He is also the associate editor of IEEE Transactions on VLSI and INTEGRATION: VLSI Journal. He has served on several NSF panels and is also the reviewer of several conferences and journals such as ICCAD, DAC, ISPD, IEEE Transactions and CAD and IEEE Transactions on VLSI. His research and teaching endeavors have received several awards including the Outstanding PhD Dissertation Award from the CS Department of UCLA, Best Paper Award from ISPD 2007, George Corcoran Memorial Teaching Award from the ECE Department of University of Maryland.