I. Introduction
In today's automotive ICs, the functional safety is of paramount importance. As defined in ISO-26262 [1], it targets “the absence of unreasonable risk due to hazards caused by malfunctioning behaviors of E/E systems”. To achieve this goal, safety mechanisms are often employed in the system to minimize the safety risk in case a system failure does occur. In the meantime, a nearly zero failure rate is required over a designated lifetime span for each safety-critical electronic component. This grand objective implies that, not only a “nearly zero defect level” needs to be targeted during the manufacturing test, various fault tolerance and online error masking techniques are inevitable to cope with sporadic soft errors induced by various sources, such as radiation, or PVTA (Process, Voltage, Temperature, and Ageing) effects. Moreover, some automotive ICs supporting self-driving could have a wide-ranging power consuming behavior due to the need of online high- performance computing, and thus there may be a greater chance to induce significant power glitches from time to time.