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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Conventional neuro-computing architectures and artificial neural networkshave often been developed with no or loose connections to neuroscience. As aconsequence, they have largely ignored key features of biological neuralprocessing systems, such as their extremely low-power consumption features ortheir ability to carry out robust and efficient computation using massivelyparallel arrays of limited precision, highly variable, and unreliablecomponents. Recent developments in nano-technologies are making availableextremely compact and low-power, but also variable and unreliable solid-statedevices that can potentially extend the offerings of availing CMOStechnologies. In particular, memristors are regarded as a promising solutionfor modeling key features of biological synapses due to their nanoscaledimensions, their capacity to store multiple bits of information per elementand the low energy required to write distinct states. In this paper, we firstreview the neuro- and neuromorphic-computing approaches that can best exploitthe properties of memristor and-scale devices, and then propose a novel hybridmemristor-CMOS neuromorphic circuit which represents a radical departure fromconventional neuro-computing approaches, as it uses memristors to directlyemulate the biophysics and temporal dynamics of real synapses. We point out thedifferences between the use of memristors in conventional neuro-computingarchitectures and the hybrid memristor-CMOS circuit proposed, and argue howthis circuit represents an ideal building block for implementing brain-inspiredprobabilistic computing paradigms that are robust to variability andfault-tolerant by design.

Integration of nanoscale memristor synapses in neuromorphic computing architectures