• DocumentCode
    3241837
  • Title

    Impact of nMOS/pMOS gate length correlation on the accuracy of statistical modeling

  • Author

    Komatsubara, Hirotaka ; Miura, Noriyuki ; Hayashi, Hirokazu ; Mochizuki, Marie ; Rukuda, K. ; Yajima, Tsukasa ; Yoshida, Masahiro ; Wakayama, Keiichi ; Nishitani, Akito ; Nagatomo, Yoshiki

  • Author_Institution
    Oki Electr. Ind. Co. Ltd., Tokyo, Japan
  • fYear
    2001
  • fDate
    2001
  • Firstpage
    47
  • Lastpage
    50
  • Abstract
    In LSI design of modern scaled-down technologies, the statistical model is used to take process variation effects into account. In order to predict device fluctuations in early stages of device development, the TCAD approach of statistical modeling has already been introduced in some reports. But from statistical modelling points of view, prediction accuracy of such TCAD approaches has not been sufficiently discussed. In this paper, statistical distribution of factory-manufactured devices is compared carefully to the prediction of TCAD. It is revealed that there is no serious accuracy problem in statistical model prediction even with economically calibrated TCAD, On the other hand, consideration of nMOS and pMOS gate length correlation is found to be critical to the prediction
  • Keywords
    MOS integrated circuits; integrated circuit manufacture; large scale integration; semiconductor process modelling; statistical analysis; technology CAD (electronics); LSI design; TCAD; factory-manufactured devices; gate length correlation; prediction accuracy; process variation effects; scaled-down technologies; statistical modeling; Calibration; Economic forecasting; Fluctuations; Intrusion detection; Large scale integration; MOS devices; Manufacturing; Predictive models; Production facilities; Threshold voltage;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Statistical Methodology, IEEE International Workshop on, 2001 6yh.
  • Conference_Location
    Kyoto
  • Print_ISBN
    0-7803-6688-3
  • Type

    conf

  • DOI
    10.1109/IWSTM.2001.933824
  • Filename
    933824