DocumentCode :
2996980
Title :
A Power and Cluster-Aware Technology Mapping and Clustering Scheme for Dual-VT FPGAs
Author :
Loke, Wei Ting ; Ha, Yajun ; Zhao, Wenfeng
Author_Institution :
Xilinx Asia Pacific, Singapore, Singapore
fYear :
2012
fDate :
21-25 May 2012
Firstpage :
221
Lastpage :
226
Abstract :
In this paper, we present a technology mapping and clustering tool for leakage power reduction in FPGAs with programmable, dual-VT logic blocks. The use of Reverse Back Bias (RBB) circuit techniques is recognized as one of the more promising strategies in mitigating leakage power, a critical problem in circuits deploying deep submicron process technologies. FPGAs with the ability to adjust fabric VT through RBB offer the potential of reducing leakage power with minimal or no sacrifice to circuit speed. Today, Altera´s Stratix line of FPGAs deploy a similar strategy, but with optimizations limited to the post-P&R stage. We present a novel two-stage technology mapping (RBBMap) and logic block packing (RBBPack) tool that is free from clustering constraints limiting the post-P&R method, and moves RBB optimizations upwards to the technology mapping level. Using the baseline technology mapping tool Emap, our tools generate an average of 70.95% savings in logic block leakage power and 28.30% savings in total energy consumption.
Keywords :
field programmable gate arrays; optimisation; pattern clustering; Alters Stratix line; Emap; RBB circuit techniques; RBB optimizations; RBBMap tool; RBBPack tool; baseline technology mapping tool; cluster-aware technology mapping; clustering tool; deep submicron process technologies; dual FPGA; energy consumption; leakage power reduction; logic block packing tool; post-P&R method; power technology mapping; programmable dual logic blocks; reverse back bias circuit techniques; two-stage technology mapping tool; Delay; Field programmable gate arrays; Optimization; Routing; Switches; Table lookup; Dual-VT; EDA; FPGA; Programmable-VT; Reverse Back Bias; Technology Mapping;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel and Distributed Processing Symposium Workshops & PhD Forum (IPDPSW), 2012 IEEE 26th International
Conference_Location :
Shanghai
Print_ISBN :
978-1-4673-0974-5
Type :
conf
DOI :
10.1109/IPDPSW.2012.23
Filename :
6270642
Link To Document :
بازگشت