A Two-Order Increase in Robustness of Partial Redundancy Under a Radiation Stress Test by Using SDC Prediction

Partial redundancy is a method to address errors from single event effects (SEEs) on critical data while leaving less important data unprotected for energy consumption trade-offs. Under a low SEE rate, the method can provide a good cost-effective fault tolerance, while many silent data corruptions (SDCs) may occur under a high fault rate due to an incomplete fault coverage. This paper proposes a system-level approach to additionally covering SDCs in a partial redundancy by a light-weighted error prediction. Our results from a simulation under a stress radiation test condition show that with an average 8% cost in energy consumption, we can reduce the SDC rate from 12% to 0.37%, for the work loads that we studied.