▎ 摘　　要

NOVELTY - The circuit has a neuron element (210) provided with a global floating gate (GFG) that is charged by voltage output from a set of synaptic elements (220, 230, 240). An electrode receives an input, and outputs the voltage to the GFG of a neuron device. A second electrode receives the input from the neuron element. A diode transmits the spike voltage to a winner takes all (WTA) bus. The GFG is made of metal material and graphene. A floating gate is arranged on a substrate. An insulating film (330) covers the floating gate, and a semiconductor channel is formed on an insulating layer. USE - Spiking neural network circuit for increasing or decreasing electrical resistance according to spike timing dependent plasticity characteristics. ADVANTAGE - The high integration and low power consumption of the spiking neural network circuit are achieved. DESCRIPTION OF DRAWING(S) - The drawing shows a structural diagram of the spiking neural network circuit. Neuron element (210) Synaptic elements (220, 230, 240) Floating gate (320) Tunneling insulating film (330) Channel layer (340)