Soft errors in SRAM-FPGAs: A comparison of two complementary approaches

As SRAM-based FPGAs are introduced in safety- or mission-critical applications, the availability of suitable electronic design automations (EDA) tools for predicting systems dependability becomes mandatory for designers. Nowadays designers can opt either for workload-independent EDA tools, which provide information about system´s dependability disregarding the workload the system is supposed to elaborate when deployed in the mission, or workload-dependant approaches. In this paper, we compare two tools for predicting the effects of soft errors in circuits implemented using SRAM-based FPGAs, a workload-independent one (STAR) and a workload-dependant one (FLIPPER). Experimental results show that the two tools are complementary and can be used fruitfully for obtaining accurate predictions.