Soft NMR: Exploiting statistics for energy-efficiency

Eric P. Kim, Rami A. Abdallah, Naresh R. Shanbhag

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Achieving energy-efficiency in nanoscale CMOS process technologies is made challenging dne to the presence of process, temperature and voltage variations. In this paper, we present soft N-modular redundancy (soft NMR) that consciously exploits statistics of errors due to these nanoscale artifacts in order to design robust and energy-efficient systems. In contrast conventional NMR, soft NMR employs estimation and detection techniques in the voter. We compare NMR and soft NMR in the design of an energy-efficient and robust discrete cosine transform (DCT) image coder. Simulations in a commercial 45 nm, 1 .2V, CMOS process show that soft triple-MR (TMR) provides lOx improvement in robustness and 13% power savings over TMR a peak signal-to-noise ratio (PSNR) of 20dB. In addition, soft dual-MR (DMR) provides 2 x improvement in robustness and 35% power savings over TMR at a PSNR of 20dB.

Original languageEnglish (US)
Title of host publication2009 International Symposium on System-on-Chip - Proceedings, SoC 2009
Pages52-55
Number of pages4
DOIs
StatePublished - Dec 1 2009
Event2009 International Symposium on System-on-Chip, SoC 2009 - Tampere, Finland
Duration: Oct 5 2009Oct 7 2009

Publication series

Name2009 International Symposium on System-on-Chip - Proceedings, SoC 2009

Other

Other2009 International Symposium on System-on-Chip, SoC 2009
Country/TerritoryFinland
CityTampere
Period10/5/0910/7/09

ASJC Scopus subject areas

  • Hardware and Architecture
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Soft NMR: Exploiting statistics for energy-efficiency'. Together they form a unique fingerprint.

Cite this