• DocumentCode
    3543359
  • Title

    Switching energy efficiency optimization for advanced CPU thanks to UTBB technology

  • Author

    Arnaud, F. ; Planes, N. ; Weber, Olivier ; Barral, V. ; Haendler, S. ; Flatresse, Philippe ; Nyer, Frederic

  • Author_Institution
    STMicroelectron., Crolles, France
  • fYear
    2012
  • fDate
    10-13 Dec. 2012
  • Abstract
    This paper presents the superior performance of UTBB (Ultra-Thin Box and Body) technology for providing high speed at low voltage. We evidence the transistor capability to sustain full forward-body-biasing solution thanks to a planar back-side gate scheme. Silicon measurements on low complexity circuits show that the dynamic power consumption can be reduced by 90% without any speed degradation by simply selecting the appropriate power supply and body bias couple (Vdd; Vbb). A simple switching energy efficiency model is then proposed allowing the (Vdd; Vbb) couple prediction reaching the minimum energy point. Finally, we demonstrate on a full CPU Core implementation with UTBB a total power reduction of -30% and a +40% energy efficiency at identical speed with respect to bulk technology thanks to back side gate biasing efficiency.
  • Keywords
    MOSFET; circuit complexity; elemental semiconductors; energy conservation; multiprocessing systems; optimisation; power supplies to apparatus; silicon; switching circuits; Si; UTBB technology; advanced CPU; back side gate biasing efficiency; body bias couple; bulk technology; couple prediction; dynamic power consumption; full CPU core implementation; full forward-body-biasing solution; low complexity circuits; minimum energy point; planar back-side gate scheme; power supply; speed degradation; switching energy efficiency model; switching energy efficiency optimization; transistor capability; ultra-thin box and body technology; Delay; Energy efficiency; Logic gates; Silicon; Switches; Transistors;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Electron Devices Meeting (IEDM), 2012 IEEE International
  • Conference_Location
    San Francisco, CA
  • ISSN
    0163-1918
  • Print_ISBN
    978-1-4673-4872-0
  • Electronic_ISBN
    0163-1918
  • Type

    conf

  • DOI
    10.1109/IEDM.2012.6478970
  • Filename
    6478970