keywords:
remainig useful life
dram
aging
This paper addresses the estimation of the remaining useful life (RUL) for second-hand integrated circuits (ICs) that experience performance degradation due to aging. While the RUL can be estimated based on curve fitting method, such as by using a fractional power function of operation time, the estimation accuracy can be limited under performance variation. This paper presents a machine learning approach on the statistical performance data of an IC over its lifespan.
