Cost-Effective Reliability Screening, Binning, and In-Field Adaptation

点击翻译按钮获取中文摘要

In the semiconductor industry, one way of measuring reliability is the expected lifetime of a product. As semiconductor manufacturing technologies continue to advance, ensuring reliability has become one of the major challenges in the industry. The ability to ensure reliability is essential for the success of a product line as well as the success of the industry. This project proposes to develop a first cost-effective solution for screening unreliable parts before product shipment to the customer. The project proposes to develop software tools and methodologies for predicting the expected life time of a part. Further, the project proposes to develop a second cost-effective solution for continuously monitoring and improving reliability while a part is being used in the field. Additional software tools and methodologies will be developed for building models for such continuous monitoring. These monitoring models are then used for extending the life time of a part by automatic adjusting the operational parameters such as speed and voltage of the part based on its current health condition. The research is integrated with educational activities to develop course and tutorial materials for broad impact, a state-of-the-art laboratory for education, and a research program to attract undergraduate and underrepresented students. The research strives to achieve a comprehensive understanding of state-of-the-art industrial practices and to accomplish multidisciplinary studies merging knowledge from reliability physics, semiconductor testing, and data mining. Knowledge discovered through this research will provide the industry with a clear direction on how to cope with the present and future reliability challenges.