Life prediction methodology of system-in-package based on physics of failure

With the dramatic development of microelectronics technology, System-in-Package (SiP) becomes a brand-new direction for the More than Moore's law. In order to satisfy the demand of small-size, multi-function and high-performance, complex structures and variable materials are applied in SiPs, which introduce many reliability problems. To implement reliability qualification and health assessment, a life prediction methodology of SiP based on physics-of-failure (PoF) is studied in conjunction with simplified life cycle profile. In this paper, typical structures of SiPs, such as dies, components, interconnects are evaluated. And related PoF mechanisms, such as time dependent dielectric breakdown, electro-migration, die attach fatigue, thermal cyclic fatigue and etc., are considered. The inputs of the methodology contain hardware information and lifecycle profile. The hardware information of SiPs includes materials types and structures size. Lifecycle profile provides environmental conditions that the SiPs should experience. Based on these inputs, thermal distributions and stress-strain distributions of the SiP are analyzed by finite element analysis (FEA) tools. With the utilization of PoF models, lifetime matrix of the SiP is obtained. The output of the methodology is the lifetime matrix to predict lifetime of the SiP. Finally, a case study is done to guide engineering applications.