Mitigating Process Variability and Soft Errors at Circuit-Level for Finfets

This book evaluates the influence of process variations (e.g. work-function fluctuations) and radiation-induced soft errors in a set of logic cells using FinFET technology, considering the 7nm technological node as a case study. Moreover, for accurate soft error estimation, the authors adopt a radiation event generator tool (MUSCA SEP3), which deals both with layout features and electrical properties of devices. The authors also explore four circuit-level techniques (e.g. transistor reordering, decoupling cells, Schmitt Trigger, and sleep transistor) as alternatives to attenuate the unwanted effects on FinFET logic cells. This book also evaluates the mitigation tendency when different levels of process variation, transistor sizing, and radiation particle characteristics are applied in the design. An overall comparison of all methods addressed by this work is provided allowing to trace a trade-off between the reliability gains and the design penalties of each approach regarding the area, performance, power consumption, single event transient (SET) pulse width, and SET cross-section.