Approximate computing is a low power technique for a class of increasingly prevalent applications, including image/voice/haptic processing, video decoding, recognition and probabilistic inference. By intentionally allowing occasional small errors, the underlying hardware for these applications can be implemented with improved energy-efficiency. Existing approximate computing techniques are focused on either component or architecture level. This talk will introduce two techniques on how to build approximate sub-systems using approximate components. The first is a general technique for scheduling and binding in high level synthesis. The second is specific for JPEG decoder circuits. The simulation results from both the techniques confirm the importance of making good use of approximate circuits.
Jiang Hu received the B. S. degree in optical engineering from Zhejiang University, China, in 1990, the M. S. degree in physics in 1997, and the Ph. D. degree in electrical engineering from the University of Minnesota in 2001. He has been with IBM Microelectronics from January 2001 to June 2002. Currently he is a professor in the Department of Electrical and Computer Engineering at the Texas A&M University. His research interests include VLSI circuit optimization, chip power management, approximate computing and hardware security. He received a best paper award at the ACM/IEEE Design Automation Conference in 2001, an IBM Invention Achievement Award in 2003 and a best paper award at the IEEE/ACM International Conference on Computer-Aided Design in 2011. He has served as General Chair and Technical Program Chair for the ACM International Symposium on Physical Design, and associated editor for IEEE Transactions on CAD and ACM Transactions on Design Automation of Electronic Systems. He is named IEEE Fellow in 2016.