Brain-inspired computing employs devices and architectures that emulate biological functions for more adaptive and energy-efficient systems. Oscillatory neural networks (ONNs) are an alternative approach in emulating biological functions of the human brain and are suitable for solving large and complex associative problems. In this work, we investigate the dynamics of coupled oscillators to implement such ONNs. By harnessing the complex dynamics of coupled oscillatory systems, we forge a novel computation model--information is encoded in the phase of oscillations. Coupled interconnected oscillators can exhibit various behaviors due to the strength of the coupling. In this article, we present a novel method based on subharmonic injection locking (SHIL) for controlling the oscillatory states of coupled oscillators that allow them to lock in frequency with distinct phase differences. Circuit-level simulation results indicate SHIL effectiveness and its applicability to large-scale oscillatory networks for pattern recognition.

Three SDRAMs from the same manufacturer with technology node sizes 110, 72, and 63 nm, were investigated under proton irradiation and using scanning electron microscopy (SEM). The radiation-induced faults were characterized and compared between the different part types. The devices under test (DUT) were irradiated with protons and experienced single event effects (SEE) in the form of stuck bits and single bit upsets (SBU). Analysis of the data retention times of bits which had SBU and were stuck during irradiation, showed similar patterns of retention time degradation, suggesting that the SBUs and stuck bits in all three part types could be induced by the same mechanism. Detailed data retention time analyses were also performed before and after irradiation to investigate the evolution of data retention times after irradiation, and after periods of annealing. The largest radiation-induced retention time losses were found to anneal, but the bits least affected directly after irradiation experienced decreasing data retention time as a function of annealing time. SEM imaging showed differences in the memory cell structure between the tested part types. The largest node size device was the most sensitive to the radiation, both for SEE and cumulative radiation effects.

IMPLANTATION MATÉRIELLE DE SYSTÈMES COMPLEXES : TRADUCTION AUTOMATIQUE D'UN RÉSEAU DE PETRI NON AUTONOME EN COMPOSANTS VHDL