Kinematic model of magnetic-domain-wall motion for fast, high-accuracy simulations
Phys. Rev. Applied 24, 034020 – Published 8 September, 2025
DOI: https://doi.org/10.1103/v9tg-9l9r
Abstract
This paper is a contribution to the Physical Review Applied collection titled Physics-Inspired Computing.
Domain-wall (DW) devices have garnered recent interest for diverse applications including memory, logic, and neuromorphic primitives; fast and accurate device models are therefore imperative for the design and verification of large-scale systems. Existing models of DW motion are suboptimal for the design of large-scale systems as they either overconsume compute resources with physics-heavy equations or oversimplify the physics, drastically reducing model accuracy. We propose a DW model inspired by the phenomenological similarities between the motions of a DW and a classical object being acted on by forces such as air resistance and static friction. Our proposed phenomenological model predicts DW motion within 1.1% on average and is 4000 times faster than micromagnetic simulations. In addition, our model is seven times faster than existing collective-coordinate models and 14 times more accurate than existing hyper-reduced models, making it an essential tool for large-scale DW circuit design and simulation. The model is publicly available, along with scripts that can automatically extract model parameters from user-provided simulation or experimental data, allowing the user to extend the model to a wider range of micromagnetic parameters.
Physics Subject Headings (PhySH)
Collections
This article appears in the following collection:
Collection on Physics-Inspired Computing
Physical Review Applied is pleased to present a Collection on Physics-Inspired Computing, highlighting the rapidly evolving field of energy-efficient computing techniques, from hardware technologies to algorithms, where physics inspiration serves as the crucial link. Contributions to this collection will be published throughout 2025. This Collection is being curated by Guest Editors Kerem Camsari and Supriyo Datta.
