Towards Quality-Driven Approximate Software Generation for Accurate Hardware: Work-in-Progress

Jorge Castro-Godinez, Muhammad Shafique, Jorg Henkel

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

Abstract

Many existing processor-based systems, especially using off-the-shelf components, cannot afford hardware modifications to embrace different approximate computing techniques proposed by the research community. In that case, it is mainly at the software level where error resiliency can be exploited efficiently. Although a multitude of approximate techniques can be applied at software level, they have been presented in isolation and little has been done to report their combined usage on different types of applications amenable to approximations. We present here AxSWGen, an automated quality-driven methodology to jointly explore and apply multiple approximate techniques to error-tolerant sections of applications. AxSWGen is implemented using LLVM compiler infrastructure. We present results of automated approximate software generated with AxSWGen and executed on a RISC-V processor (SiFive HiFive1 board), achieving up to 50% energy reduction for a 5% image degradation for an approximate Gaussian filter.

Original languageEnglish (US)
Title of host publicationProceedings of the 2020 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES 2020
EditorsTulika Mitra, Andreas Gerstlauer
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages12-14
Number of pages3
ISBN (Electronic)9781728191928
DOIs
StatePublished - Sep 20 2020
Event2020 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES 2020 - Virtual, Singapore, Singapore
Duration: Sep 20 2020Sep 25 2020

Publication series

NameProceedings of the 2020 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES 2020

Conference

Conference2020 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES 2020
CountrySingapore
CityVirtual, Singapore
Period9/20/209/25/20

Keywords

  • component
  • formatting
  • insert
  • style
  • styling

ASJC Scopus subject areas

  • Hardware and Architecture
  • Safety, Risk, Reliability and Quality

Fingerprint Dive into the research topics of 'Towards Quality-Driven Approximate Software Generation for Accurate Hardware: Work-in-Progress'. Together they form a unique fingerprint.

Cite this