Intrinsic Spike-Timing-Dependent Plasticity in Stochastic Magnetic Tunnel Junctions Mediated by Heat Dynamics

The quest for highly efficient cognitive computing has led to extensive research interest in the field of neuromorphic computing. Neuromorphic computing aims to mimic the behavior of biological neurons and synapses using solid-state devices and circuits. Among various approaches, emerging nonvolatile memory technologies are of special interest for mimicking neuro-synaptic behavior. These devices allow the mapping of the rich dynamics of biological neurons and synapses onto their intrinsic device physics. In this letter, we focus on spike-timing-dependent plasticity (STDP) behavior of biological synapses and present a method to implement the STDP behavior in magnetic tunnel junction (MTJ) devices. Specifically, we exploit the time-dependent heat dynamics and the response of an MTJ to the instantaneous temperature to imitate the STDP behavior. Our simulations, based on a macrospin model for magnetization dynamics, show that STDP can be imitated in stochastic MTJs by applying simple voltage waveforms as the spiking response of pre- and postneurons across an MTJ device.

Automated summary

The quest for highly efficient cognitive computing has sparked extensive research interest in neuromorphic computing, which mimics biological neurons and synapses using solid-state devices. Nonvolatile memory, particularly magnetic tunnel junction (MTJ) devices, are of interest for mimicking neuro-synaptic behavior. This study focuses on implementing spike-timing-dependent plasticity (STDP) behavior in MTJ devices by utilizing time-dependent heat dynamics and response to temperature changes. Simulation results show that STDP can be imitated in stochastic MTJs through simple voltage waveforms.