Z. Ji, J. F. Zhang, L. Lin, M. Duan, W. Zhang, X. Zhang, R. Gao, B. Kaczer, J. Franco, T. Schram, N. Horiguchi, S. De Gendt, G. Groeseneken
{"title":"A test-proven As-grown-Generation (A-G) model for predicting NBTI under use-bias","authors":"Z. Ji, J. F. Zhang, L. Lin, M. Duan, W. Zhang, X. Zhang, R. Gao, B. Kaczer, J. Franco, T. Schram, N. Horiguchi, S. De Gendt, G. Groeseneken","doi":"10.1109/VLSIT.2015.7223693","DOIUrl":null,"url":null,"abstract":"For the first time, we demonstrate that A-G model extracted from short Vg-accelerated stresses can predict both long term DC and AC NBTI under low and dynamic operation Vg. This is achieved by successfully separating non-saturating defects from the saturating ones, allowing reliable extraction of power exponents needed for long term prediction. Unlike R-D model, A-G model does not require solving differential equations for AC NBTI. This saves computation time significantly, especially for high-frequency that needs small time-step, and makes it readily implementable in SPICE-like simulators.","PeriodicalId":181654,"journal":{"name":"2015 Symposium on VLSI Technology (VLSI Technology)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2015-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"11","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 Symposium on VLSI Technology (VLSI Technology)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/VLSIT.2015.7223693","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 11
Abstract
For the first time, we demonstrate that A-G model extracted from short Vg-accelerated stresses can predict both long term DC and AC NBTI under low and dynamic operation Vg. This is achieved by successfully separating non-saturating defects from the saturating ones, allowing reliable extraction of power exponents needed for long term prediction. Unlike R-D model, A-G model does not require solving differential equations for AC NBTI. This saves computation time significantly, especially for high-frequency that needs small time-step, and makes it readily implementable in SPICE-like simulators.