Title :
Statistical distributions of electromigration induced failures and their impact on reliability prediction and design
Author :
Loupis, M.I. ; Avaritsiotis, J.N.
Author_Institution :
Div. of Comput. Sci., Nat. Tech. Univ. of Athens, Greece
Abstract :
In electromigration failure studies it is generally assumed that electromigration induced failures may be adequately modelled by a log normal distribution. In order to examine the applicability of the logarithmic distribution of extreme values in electromigration a novel model was created, using an object oriented and fully 2-dimensional electrical and thermal sub-models. The repeated simulations performed showed that the log extreme value distribution is a better statistical model for electromigration failures. The significance of such a modelling is particularly apparent in electromigration failure rate prediction
Keywords :
VLSI; electromigration; failure analysis; integrated circuit metallisation; integrated circuit modelling; integrated circuit reliability; reliability theory; VLSI interconnects; electrical sub-models; electromigration induced failures; failure rate prediction; log extreme value distribution; object oriented models; reliability prediction; thermal sub-models; Conductors; Electromigration; Grain size; Log-normal distribution; Object oriented modeling; Statistical analysis; Statistical distributions; Testing; Thermal resistance; Weibull distribution;
Conference_Titel :
Microelectronics, 1995. Proceedings., 1995 20th International Conference on
Print_ISBN :
0-7803-2786-1
DOI :
10.1109/ICMEL.1995.500871