The functional safety of a system or part of a system refers to the correct operation of the system in response to its input, which includes management of errors, hardware failure, and changes to operating conditions. The two types of faults that can cause a system failure and result in a violation of the functional safety goals are systematic faults and random faults.
Systematic faults arise from errors in the development or manufacturing processes. When defects appear in hardware or software, they are systematic faults. Some of the causes of systematic faults are a failure to verify intended functionality, manufacturing test escapes, or operating a device outside of a specified range. The mitigation of systematic faults is achieved by robust best practices and processes defined by safety standards.
Random faults are inherent due to silicon aging, environmental conditions, etc. Safety standards focus on detecting and managing random faults.
This chapter provides an overview of the safety mechanisms implemented in the AMD Versal™ adaptive SoC. The features are grouped into these categories:
- Single point fault detection
- Common cause failure detection
- Latent fault detection
- Isolation features
- Additional features