Predictive Analysis of Configuration Bit Criticality in SRAM-Based FPGAs. Methodology, tool, and results
Résumé
SRAM based FPGAs use has increased in the space application field in the last few years. Those devices may operate in harsh environnements such as ionizing radiations, when their application requires a high level of safety. While these FPGAs are not developed on radiation-hardened technologies, a predictive analysis of their configuration memory criticality to faults allows an radiation-hardened implementation of designs on those devices. The presentation will define our criticality classement of configuration bits, then introduce algorithms that perform the actual classification with regards to the design use of the FPGA resources, then draw the first results (AIRBUS application implemented on an AT40K ATMEL).