Professor Jason Cong Receives ACM TODAES Best Paper Award
Professor Jason Cong is a recipient of the 2012 ACM Transactions on Design Automation of Electronic Systems (TODAES) Best Paper Award for the journal entitled "Behavior-Level observability Analysis for Operation Gating in Low-Power Behavioral Synthesis," with co-authors from UCLA Bin Liu and Rupak Majumdar, and Zhiru Zhang from AutoESL Design Technologies, Inc. The award was presented on June 5, 2012 at the Design Automation Conference held in San Francisco, CA.
The journal was published on ACM TODAES in November 2010. The selection was based on three points provided by ACM TODAES:
It is the first to formally generalize the concept of ODC in the behavior level. In addition, it presents an exact and efficient algorithm to compute the behavior-level ODC in the presence of black-box operations. Traditional (or logic) ODC based analysis can hardly derive high-level ODC information (esp. across clock boundaries) in complete and efficient manners.
It is also the first to combine behavior-level ODC analysis with high-level synthesis (HLS) optimizations to identify and avoid unnecessary computations and data transfers.
The proposed ODC-based optimizations have been integrated into a cutting-edge commercial HLS system (AutoESL/Xilinx), leading to significant power savings (30%+) for the synthesized circuits.
Click here to access the full journal.