论文部分内容阅读
随着ＶＬＳＩ技术的发展，电迁移已成为集成电路最主要的失效原因之一，其可靠性评估技术也显得愈加重要。要改进该技术，不仅需要确定可靠性物理模型参数，而且要求掌握参数的统计分布特性。基于电迁移物理模型，提出一种提取参数统计分布特性的新方法。与传统方法相比，此方法不仅所需实验样品少、实验次数少，能真正得到反映样品离散性的物理模型参数的统计分布特性，而且也可用于其它失效机理（如栅氧击穿）物理模型中。
With the development of VLSI technology, electromigration has become one of the most important failure causes of integrated circuits, and its reliability evaluation technology has become even more important. To improve this technique, not only the parameters of the reliability physics model need to be determined, but also the statistical distribution of the parameters must be mastered. Based on the physical model of electromigration, a new method to extract the statistical distribution of parameters is proposed. Compared with the traditional method, this method not only requires less experimental samples and fewer experiments, but also can obtain the statistical distribution of the physical model parameters that reflect the discreteness of the sample, and can also be used for other failure mechanisms (such as gate oxygen breakdown) Model.