Neural Computing With Magnetoelectric Domain-Wall-Based Neurosynaptic Devices

Conventional von-Neumann computing models have achieved remarkable feats for the past few decades. However, they fail to deliver the required efficiency for certain basic tasks like image and speech recognitions when compared with the biological systems. As such, taking cues from the biological systems, novel computing paradigms are being explored for the efficient hardware implementations of the recognition/classification tasks. The basic building blocks of such neuromorphic systems are neurons and synapses. Toward that end, we propose a leaky-integrate-fire (LIF) neuron and a programmable non-volatile synapse using the domain-wall (DW) motion induced by the magnetoelectric effect. Due to the strong elastic pinning between the ferromagnetic DW (FM-DW) and the underlying ferroelectric DW (FE-DW), the FM-DW is dragged by the FE-DW on the application of a voltage pulse. The fact that FE materials are insulators allows for pure voltage-driven FM-DW motion, which in turn can be used to mimic the behavior of the biological spiking neurons and synapses. The voltage-driven nature of the proposed devices allows the energy-efficient operation. A detailed device to the system-level simulation framework based on the micromagnetic simulations has been developed to analyze the feasibility of the proposed neurosynaptic devices for implementing the neuromorphic systems. A key highlight of the presented work as opposed to the prior works on the DW neurons and synapses is that the proposed device can seamlessly incorporate the “controlled leaky” behavior both in neurons and synapses, leading to the improved bioplausible behavior.