English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Memristive devices represent a promising technology for building neuromorphicelectronic systems. In addition to their compactness and non-volatilityfeatures, they are characterized by computationally relevant physicalproperties, such as state-dependence, non-linear conductance changes, andintrinsic variability in both their switching threshold and conductance values,that make them ideal devices for emulating the bio-physics of real synapses. Inthis paper we present a spiking neural network architecture that supports theuse of memristive devices as synaptic elements, and propose mixed-signalanalog-digital interfacing circuits which mitigate the effect of variability intheir conductance values and exploit their variability in the switchingthreshold, for implementing stochastic learning. The effect of devicevariability is mitigated by using pairs of memristive devices configured in acomplementary push-pull mechanism and interfaced to a current-mode normalizercircuit. The stochastic learning mechanism is obtained by mapping the desiredchange in synaptic weight into a corresponding switching probability that isderived from the intrinsic stochastic behavior of memristive devices. Wedemonstrate the features of the CMOS circuits and apply the architectureproposed to a standard neural network hand-written digit classificationbenchmark based on the MNIST data-set. We evaluate the performance of theapproach proposed on this benchmark using behavioral-level spiking neuralnetwork simulation, showing both the effect of the reduction in conductancevariability produced by the current-mode normalizer circuit, and the increasein performance as a function of the number of memristive devices used in eachsynapse.

A neuromorphic systems approach to in-memory computing with non-ideal memristive devices: From mitigation to exploitation