• DocumentCode
    1996632
  • Title

    Power gating scheduling for power/ground noise reduction

  • Author

    Hailin Jiang ; Marek-Sadowska, M.

  • Author_Institution
    Dept. of Electr. & Comput. Eng., California Univ., Santa Barbara, CA
  • fYear
    2008
  • fDate
    8-13 June 2008
  • Firstpage
    980
  • Lastpage
    985
  • Abstract
    Power gating is a technique for efficiently reducing leakage power by disconnecting idle blocks from the power grid. When gated blocks are woken up, large amounts of switching currents are drawn in a short period of time that may introduce severe noise on the power delivery mesh. In this paper, we propose a GA-based approach to schedule power gating considering power/ground noise. We introduce a simulation-based method to accurately and efficiently estimate the worst case noise, taking all the current sources, inductance and decaps´ effects into consideration. We also present an incremental scheduling procedure considering the dynamic changes of decap configuration. Experimental results show that by optimally scheduling the wake-up order under time constraints, our technique can reduce noise up to 50% compared to waking gated blocks simultaneously. The quality of results depends upon the total wake-up time constraint, locations of gated blocks, current densities of gated blocks, and decap distribution.
  • Keywords
    current density; genetic algorithms; integrated circuit noise; leakage currents; power grids; power supply circuits; GA approach; current densities; decap configuration; decap distribution; gated blocks; genetic algorithms; leakage power; power gating scheduling; power grid; power/ground noise reduction; simulation-based method; wake-up order; wake-up time constraint; Circuits; Clocks; Delay; Noise reduction; Permission; Power grids; Power supplies; Sleep; Time factors; Transistors; Power gating; power supply noise; scheduling;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Design Automation Conference, 2008. DAC 2008. 45th ACM/IEEE
  • Conference_Location
    Anaheim, CA
  • ISSN
    0738-100X
  • Print_ISBN
    978-1-60558-115-6
  • Type

    conf

  • Filename
    4555962