Failure and root cause analysis for a system-on-chip: An industrial case study

Abstract

Summary form only given: This paper discusses an industrial case study on failure and root cause analysis for a system-on-chip (SoC). This SoC has high stuck-at test coverage but a single stuck-at fault causes the chip to fail which results in a high DPPM (Defect Part Per Million). We first show the complexity and issues diagnosing this failure and then cover the basic diagnosis steps localizing the defect starting from DFD (Design-For-Debug), test vector generation to screen the defect and failure analysis. A solution to avoid these type of failures is also discussed.