Graphene/MoS2/SiOx memristive synapses for linear weight update

Abstract Memristors for neuromorphic computing have gained prominence over the years implementing synapses and neurons due to their nano-scale footprint reduced complexity. Several demonstrations show two-dimensional (2D) materials as a promising platform realization of transparent, flexible, ultra-thin memristive synapses. However, unsupervised learning in spiking neural network (SNN) facilitated by linearity symmetry synaptic weight update has not been explored thoroughly using 2D platform. Here, we demonstrate that graphene/MoS 2 /SiO x /Ni exhibit ideal when subjected identical input pulses, which is essential role online training networks. The holds range pulse width, amplitude number applied pulses. Our work illustrates mechanism switching MoS -based through conductive filaments governed Poole-Frenkel emission. We synapses, integrated with leaky integrate-and-fire neuron, can control neuron efficiently. This establishes viable all-memristive SNNs.