Spintronic Neuromorphic Hardware Using Domain Wall-Based Neurons and Quantized Synapses

Abstract

In this work, we simulate the functionality of an artificial neuron and synapse using spin-orbit torque-based spintronic devices and implement a fully connected artificial neural network (ANN). These neuro-synaptic devices are emulated via transverse domain wall dynamics within a rectangular magnetic nanotrack comprised of heavy-metal (HM)/ferromagnet (FM) heterostructures. Here, the heterostructure is parameterized using Pt-based alloys for the HM layer and CoFeB for the FM layer. The ReLU activation function of the neuron has been mimicked using the domain wall (DW) motion induced by a 3 ns current pulse. The synapse is modeled using current-induced DW dynamics with a corrugated HM/FM nanotrack. Semicircular notches are symmetrically positioned along both edges of the nanotrack to serve as pinning sites. By applying 10 ns current pulses of varying current densities, we achieve controlled, stepwise DW motion characterized by temporary pauses at consecutive pinning centers. The electrical conductance of the pinned DW across various pinning points act as stable synaptic weights for our ANN. Furthermore, we observe a threshold-dependent delay effect where each depinning event is influenced by previous ones, successfully mimicking synaptic states and adaptability in neuromorphic systems. The fully connected ANN has been modeled using the conventional float32 synaptic weights for the MNIST and Fashion-MNIST datasets with an accuracy of ~97% and ~86%, respectively, which serves as a test bed for our neuromorphic simulations. To implement a sparse and low-memory-footprint ANN, we quantize the trained synaptic weights into discrete quantized levels and tested the network, which demonstrate an accuracy of ~95% and ~62% for the MNIST and Fashion-MNIST datasets, respectively.