• DocumentCode
    3193814
  • Title

    Power-aware partitioning of data converters

  • Author

    Bonanno, Alberto ; Bocca, Alberto ; Macii, Alberto ; Macii, Enrico

  • Author_Institution
    Politec. di Torino, Torino, Italy
  • fYear
    2010
  • fDate
    27-29 Sept. 2010
  • Firstpage
    358
  • Lastpage
    363
  • Abstract
    Serial data streaming, one of the most important functions in modern communication systems, is becoming more and more power consuming as bit-rate is increasing without standstill. In this work, we propose a novel technique for partitioning conventional N-bit registers in standard data converters, in order to reduce their switching activity, and therefore power consumption. The architecture here presented have a very low area overhead with respect to the standard ones for serializers and, furthermore, it allows different (i.e., custom) configurations for the partitioning. The proposed method even allows to extract idleness conditions of register banks in order to apply the well-known clock-gating technique to the circuit and thus furtherly reducing the total power consumption. This method has been applied to different data converters (i.e., serializers) in a base-band radio within an ultra low-power industrial design and the results highlight the effectiveness of the proposed technique.
  • Keywords
    data conversion; logic partitioning; power aware computing; shift registers; base band radio; clock gating technique; conventional N-bit registers; data converters; modern communication system; power aware partitioning; power consuming; power consumption; register banks; serial data streaming; standard data converter; switching activity reduction; ultra low power industrial design; Clocks; Computer architecture; Converters; Optimization; Radiation detectors; Shift registers; Data Converters; Low-power Design; Serializers; Shift Registers;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    VLSI System on Chip Conference (VLSI-SoC), 2010 18th IEEE/IFIP
  • Conference_Location
    Madrid
  • Print_ISBN
    978-1-4244-6469-2
  • Type

    conf

  • DOI
    10.1109/VLSISOC.2010.5642687
  • Filename
    5642687