Electrical Engineering Seminar and Special Problems – TST & RELIB OF ICS
3 credits
Modern, integrated circuits (ICs) and systems on chip (SoCs) fabricated in state of the art (nanoscale) manufacturing process technologies can contain tens of billions of transistor devices interconnected in extremely complex patterns spread on multiple layers of conducting (metal), semiconducting (silicon, germanium) and dielectric (silicon dioxide) materials. These ICs are prone to manufacturing defects that can manifest themselves immediately following their production or over longer periods of time while a product is deployed in the field. This course will focus on design approaches/methodologies and electronic design automation (EDA) algorithms used to effectively test ICs post their manufacturing. The topics of physical fault modeling, fault simulation, test pattern generation, design for testability, scan design, built-in self-test are among the many topics relevant to IC testing and design for testability that will be presented and discussed.
Understanding the mechanisms that affect the reliability (i.e., correct functionality and performance over time) of ICs is also critical as an increasingly large number of ICs are being deployed in critical applications like autonomous driving, medical and space applications. Mechanisms like bias temperature instability (BTI), hot carrier injection (BCI) and electromigration will be considered in depth, as well as recently developed and proposed reliability-aware design methodologies. Design or other approaches to mitigate IC aging will be studied.
Machine Learning and AI-based approaches are being increasingly used to advance the specialized fields of test and reliability. Examples of ML in test and reliability will be included in the course discussions and presentations. The course will introduce the fundamentals of IC test and reliability and some of the best-established industrial practices. Course participants will be expected to delve into some of the latest research questions and findings related to test and/or reliability. Grading for the course will be largely based on individual (and/or group) projects and presentations.