Abstract

Current-induced domain wall motion (CIDWM) shows promising prospects with low power, high density, high speed, and so on. Recent studies have demonstrated that magnetic tunnel junction (MTJ) based on CIDWM has great potential in mimicking a non-volatile artificial neuron and synapse. By combining the effect of the interfacial Dzyaloshinskii–Moriya interaction (DMI) to stabilize the Néel-type domain wall (DW) and the high-speed low-power advantages of spin orbit torque (SOT), the novel MTJ device is experimentally demonstrated with low threshold current and high propagation speed, which shows potential applications in the field of the artificial neuron and synapse. In this article, a compact model for CIDWM-MTJ based on SOT and DMI is presented. The model integrates the SOT mechanism for magnetization reversal and DW nucleation, CIDWM behaviors, and tunnel resistance theory of MTJ nanopillar. The micromagnetic simulation and the temporal evolution of the DW position are implemented by solving the Landau–Lifshitz–Gilbert (LLG) equation and the 1-D model. Based on the developed model, a hybrid CIDWM-MTJ/CMOS circuit is simulated for verification. The presented model combines the CIDWM dynamics and the MTJ magnetic dynamics, showing potential application for the simulation of all-spin artificial neural network (ANN).

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call