Today’s SoCs for automotive safety-related systems integrate numerous IP blocks. At the system level, the Hardware Software Interface (HSI) between these IP blocks needs to be verified in simulation and validated in prototype. Depending on the Automotive Safety Integrity Level (ASIL) of the application, it is necessary to ensure that any fault in a safety-related logic in hardware is detected either by hardware or software or both, before it translates into a failure. Apart from the systematic faults like specification, design, and manufacturing faults that have a deterministic cause, the random hardware faults that occur unpredictably add to the challenge of detection for resolution well within a specified fault reaction time interval.
Since the random hardware faults can occur in any safety-related part of the SoC, there is a need for IP and SoC developers to implement safety mechanisms that are internal or external to IP blocks, in order to detect, report, and control fault occurrences. This white paper explains how automotive IP developers need to implement safety standards to make safety integration easier at the SoC level.