International audienceEvaluating the sensitivity to soft-errors of integrated circuits and systems became a main issue especially if they are intended to operate in space or at high altitudes. In this paper, a new fully automated SEU fault-injection method is presented and illustrated by its application to an 8051 microcontroller. Predicted SEU error-rates are in a good agreement with results issued from radiation ground testing, thus putting in evidence the accuracy of the studied method

No abstract

Abstract-Radiation tests with 15-MeV neutrons were performed in a COTS SRAM including a new memory cell design combining SRAM cells and DRAM capacitors to determine if, as claimed, it is soft-error free and to estimate upper bounds for the cross-section. These tests led to cross-section values two orders of magnitude below those of typical CMOS SRAMs in the same technology node. MUSCA SEP3 simulations complement these results predicting that only high-energy neutrons (> 30 MeV) can provoke bit flips in the studied SRAMs. MUSCA SEP3 is also used to investigate the sensitivity of the studied SRAM to radioactive contamination and to compare it with the one of standard CMOS SRAMs. Results are useful to make predictions about the operation of this memory in environments such as avionics.

No abstract

No abstract

No abstract

The associative Hopfield memory is a form of recurrent Artificial Neural Network (ANN) that can be used in applications such as pattern recognition, noise removal, information retrieval, and combinatorial optimization problems. This paper presents the implementation of the Hopfield Neural Network (HNN) parallel architecture on a SRAM-based FPGA. The main advantage of the proposed implementation is its high performance and cost effectiveness: it requires O(1) multiplications and O(log N) additions, whereas most others require O(N) multiplications and O(N) additions.

International audienceA method and the corresponding platform devoted to operational SEE-rate prediction are presented and illustrated by experimental results. Predicted error-rates are in well agreement with results issued from the activation of an SRAM platform, in 90 nm technology node, on board stratospheric balloons flights. Direct ionization of protons is investigated for a 65 SRAM memory virtually boarded on the balloon flight