Microelectronic reliability predictions derived from component defect densities

Abstract

A physics-of-failure approach to reliability prediction for integrated circuits is discussed. The analysis described is based upon the expectation that no integrated circuit can ever be free of imperfections and the assumption that both microscopic (point) defects and macroscopic flaws play influential roles in determining IC reliability. It is demonstrated that the microscopic defects can be directly implicated in gradual degradation over time via analyses related to those used in modeling a variety of solid-state phenomena.<>