Reducing computation overhead of flash translation layer with hash

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

NAND flash memory does not support the overwrite operation and thus it deploys the flash translation layer which performs the out-place update. When designing a flash translation layer, memory consumption, computation complexity, and garbage collection overhead should be low. However, the representative sector mapping scheme of flash translation layer, FAST causes a huge computation complexity even though it minimizes the memory consumption and the garbage collection overhead. This paper presents a hashed page table for the FAST scheme to reduce the computation overhead. The simulation result shows that the presented method contributes to reduce the computation overhead considerably.

Original languageEnglish
Title of host publicationProceedings of the 9th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS '10
Pages126-129
Number of pages4
StatePublished - 2010
Event9th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS '10 - Hangzhou, China
Duration: 11 Apr 201013 Apr 2010

Publication series

NameProceedings of the 9th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS '10

Conference

Conference9th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS '10
Country/TerritoryChina
CityHangzhou
Period11/04/1013/04/10

Keywords

  • Flash translation layer
  • Hash
  • Log block
  • NAND flash memory
  • Sector mapping

Fingerprint

Dive into the research topics of 'Reducing computation overhead of flash translation layer with hash'. Together they form a unique fingerprint.

Cite this