Voltage-Controlled Domain Wall Motion-Based Neuron and Stochastic Magnetic Tunnel Junction Synapse for Neuromorphic Computing Applications

This work discusses the proposal of a spintronic neuromorphic system with spin orbit torque-driven domain wall motion (DWM)-based neurons and synapses. We propose voltage-controlled magnetic anisotropy DWM-based tunnel junction (MTJ) neuron. investigate how electric field at gate (pinning site), generated by voltage signals from pre-neurons, modulates DWM, which reflects in nonlinear switching behavior neuron magnetization. For implementation synaptic weights, we 3-terminal MTJ stochastic DWM free layer. incorporate intrinsic pinning effects creating triangular notches on sides The thermal noise device lead to DWM. control this stochasticity torque is shown realize potentiation depression weight. micromagnetics transport studies synapses are carried out developing coupled micromagnetic non-equilibrium Green’s function ( MuMag-NEGF ) model. minimization writing current pulsewidth leveraging demagnetization energy also presented. Finally, discuss digit recognition proposed using spike time-dependent algorithm.