No abstract

No abstract

International audienceThis work revisits the stochastic computing paradigm as a way to implement architectures dedicated to probabilistic inference. In general, it is assumed the operation over stochastic bit streams is robust with respect to radiation transient events effects. Moreover, it can be expected that leveraging the stochastic computing paradigm to implement probabilistic computations such as Bayesian inference implemented in hardware, could yield an increased resilience to radiation effects comparatively to deterministic procedures. However, the practical assessment of the robustness against radiation is mandatory before considering Stochastic Bayesian Machines (SBMs) in hazardous environments. Results of fault injection campaigns at RTL level provide the first evidences of the intrinsic robustness of SBMs with respect to SEUs and SETs

No abstract

No abstract

No abstract

No abstract

The use of FPGA's for implementing fault-tolerant systems (FTS) has been widely discussed. Many FTS's have been proposed in this context and TMR is by far the most used architecture. However, those implementations have to count on the memory configuration's integrity of those FPGA's, since all the TMR's circuitry is stored into it. In fact, radiation or even electromagnetic noise can disturb the content of the configuration memory, with disastrous results for the system. In this paper we propose a way of dealing with this problem by using the FPGA's CRC as its signature. In case of an error derived from a kind of fault mentioned above, that signature will change. By voting those signatures in TRM architectures we can not only detect the faults, but we can recover from them by copying the memory configuration of a faultless FPGA into a faulty one. We discuss the difficulties of implementing this technique and the workarounds used to get over those difficulties. Finally we implement an experiment to validate the idea.