• DocumentCode
    2546460
  • Title

    Fast Decoupling Capacitor Budgeting for Power/Ground Network Using Random Walk Approach

  • Author

    Kang, Le ; Cai, Yici ; Zou, Yi ; Shi, Jin ; Hong, Xianlong ; Tan, Sheldon X D

  • Author_Institution
    Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing
  • fYear
    2007
  • fDate
    23-26 Jan. 2007
  • Firstpage
    751
  • Lastpage
    756
  • Abstract
    This paper proposes a fast and practical decoupling capacitor (decap) budgeting algorithm to optimize the power ground (P/G) network design. The new method adopts a modified random walk process to partition the circuit. Then, by utilizing the isolation property of decaps, this new method avoids solving the large nonlinear programming problem in traditional decap optimization process. Also, this method integrates leakage currents optimization algorithm using a refined leakage model. Experimental results demonstrate that our proposed method achieves approximate a 10times speed up over the heuristic method based on sensitivity and only about 6% decap area deviation from the optimal budget using the programming method.
  • Keywords
    circuit optimisation; integrated circuit design; leakage currents; logic partitioning; random processes; decap budgeting algorithm; decap optimization process; decoupling capacitor budgeting algorithm; heuristic method; isolation property; leakage currents optimization algorithm; power ground network design; programming method; random walk approach; refined leakage model; Circuit simulation; Computer science; Isolation technology; Noise robustness; Optimization methods; Partitioning algorithms; Power capacitors; Power system reliability; Voltage fluctuations; Wire;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Design Automation Conference, 2007. ASP-DAC '07. Asia and South Pacific
  • Conference_Location
    Yokohama
  • Print_ISBN
    1-4244-0629-3
  • Electronic_ISBN
    1-4244-0630-7
  • Type

    conf

  • DOI
    10.1109/ASPDAC.2007.358079
  • Filename
    4196125