Effective instruction fetch stage design for 16-bit instruction set architecture

Kim Areum, Joong Hwang Seok, Wook Kim Seon

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

Abstract

The 16-bit instruction set architecture has merits in terms of code size reduction and instruction cache efficiency. But we have taken the advantages at the cost of performance due to lack of an available expression space for a long immediate value, a three-address mode and so on. This paper presents a new instruction coalescing technique named as move folding to remove redundant move instructions caused by the limitation of the 16-bit instruction set. We prove effectiveness of the technique by implementing it on a commercial microprocessor. The proposed move folding technique improves speedup of 5% on average and up to 18% at the cost of 4.3% hardware complexity increment.

Original languageEnglish
Title of host publicationProceedings - 8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008
Pages563-568
Number of pages6
DOIs
Publication statusPublished - 2008
Event8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008 - Sydney, Australia
Duration: 2008 Jul 82008 Jul 11

Publication series

NameProceedings - 8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008

Other

Other8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008
Country/TerritoryAustralia
CitySydney
Period08/7/808/7/11

ASJC Scopus subject areas

  • Computer Science(all)
  • Computer Graphics and Computer-Aided Design
  • Information Systems
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Effective instruction fetch stage design for 16-bit instruction set architecture'. Together they form a unique fingerprint.

Cite this