Alyssa Bonnoit Carnegie Mellon University

Integrated Dynamic Voltage/Frequency Scaling and Adaptive Body Biasing using Test-time Voltage Selection

Adaptive body biasing (ABB) is a promising technique for addressing increasing process variability, but it also provides new opportunities for reducing power when combined with dynamic voltage/frequency scaling (DVFS). Limitations of existing ABB/DVFS proposals are explored, and a new scheme, test-time voltage selection (TTVS), is presented. By delaying the mapping between frequency and supply voltage until test, variability information can be incorporated into the supply voltage selection process. For a 16-core chip-multiprocessor implemented in a high-performance 22 nm technology, TTVS results in 18% power savings over independent ABB/DVFS and 11% power savings over the best of several previously proposed ABB/DVFS schemes.

Bio:

Alyssa Bonnoit is a PhD candidate advised by Professor Larry Pileggi in the Department of Electrical and Computer Engineering at Carnegie Mellon University. She received her B.S. in electrical engineering from Swarthmore College in 2003, and worked at IBM from 2003 through 2006. Her research is presently focused on variation-aware design of scaled CMOS technologies.