• DocumentCode
    2343513
  • Title

    Performance Analysis of Variable Number of Sets in Fully Associative Cache

  • Author

    Subha, S.

  • Author_Institution
    Dept. of Comput. Eng., Santa Clara Univ., Santa Clara, CA, USA
  • fYear
    2009
  • fDate
    2-4 April 2009
  • Firstpage
    295
  • Lastpage
    298
  • Abstract
    Energy consumption in caches depends on the number of sets/ways that are active during program execution. A fully associative cache with modified address translation using XOR function has been proposed in literature. This paper compares the performance of this model with direct, set associative, fully associative memory of same size and analyzes the energy consumed in the proposed model. Expressions for the average memory access time for the proposed model are stated. Energy consumed in the proposed model is compared with the direct, set associative and fully associative cache of same size and conditions for outperforming them are derived. Simulations are done on SPEC 2000 benchmarks. The performance with respect to average memory access time is found to be equal to direct, set associative, fully associative memory of same size for the chosen parameters. The energy consumed is comparable with a set associative cache of same size and ways. An improvement in the energy consumed of 99% is seen in the case of the fully associative memory of same size.
  • Keywords
    content-addressable storage; performance evaluation; power aware computing; storage management; SPEC 2000; XOR function; associative cache; associative memory; energy consumption; performance analysis; program execution; Analytical models; Associative memory; Cache memory; Energy consumption; Mathematical analysis; Performance analysis; Power engineering and energy; Time measurement; Average Memory Access Time; Fully Associative Memory;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computing, Engineering and Information, 2009. ICC '09. International Conference on
  • Conference_Location
    Fullerton, CA
  • Print_ISBN
    978-0-7695-3538-8
  • Type

    conf

  • DOI
    10.1109/ICC.2009.36
  • Filename
    5328160