Skip to main navigation Skip to search Skip to main content

A high speed and low power content-addressable memory(CAM) using pipelined scheme

  • SUNY Buffalo

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

5 Scopus citations

Abstract

This paper presents a novel technique to design high performance Content-addressable memories(CAMs), with lower power and latency as compared to other similar structures. The first technique is to pipeline the search operation by distributing single matching operation into several segments for different search-line registers. Speed is improved significantly since four search-line registers are comparing in parallel. Meanwhile, by disabling the subsequent segments, the power consumption is also reduced. The second technique is to improve the speed further by using multi-bank search data registers structure. The experimental results show that up to 37.32% power savings can be obtained and 90.79% time can be shrieked as compared to conventional NOR-type CAM design.

Original languageEnglish
Title of host publicationProceedings - 28th IEEE International System on Chip Conference, SOCC 2015
EditorsThomas Buchner, Danella Zhao, Karan Bhatia, Ramalingam Sridhar
PublisherIEEE Computer Society
Pages345-349
Number of pages5
ISBN (Electronic)9781467390934
DOIs
StatePublished - Feb 12 2016
Event28th IEEE International System on Chip Conference, SOCC 2015 - Beijing, China
Duration: Sep 8 2015Sep 11 2015

Publication series

NameInternational System on Chip Conference
Volume2016-February
ISSN (Print)2164-1676
ISSN (Electronic)2164-1706

Conference

Conference28th IEEE International System on Chip Conference, SOCC 2015
Country/TerritoryChina
CityBeijing
Period09/8/1509/11/15

Keywords

  • Content-addressable memory (CAM)
  • high speed
  • low power
  • pipeline

Fingerprint

Dive into the research topics of 'A high speed and low power content-addressable memory(CAM) using pipelined scheme'. Together they form a unique fingerprint.

Cite this