EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Harnessing Time‐Dependent Magnetic Texture Dynamics via Spin‐Orbit Torque for Physics‐Enhanced Neuromorphic Computing

Yifan Zhang, Yu Li, Huai Lin, Xinying Wang, Long Liu, Guoliang Xing, Di Wang, Zhihao Zhao, Zhipeng Guo, Junbo Yang, Jiebin Niu, Yan Sun und 2 weitere

Advanced Science · 2026-02 · 13 (12)

DOI: 10.1002/advs.202513946 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Science, online erschienen am 10.11.2025. Verfasst von Yifan Zhang und 13 weiteren Autor:innen. Beteiligte Einrichtung: State Key Laboratory of Fabrication Technologies for Integrated Circuits, Institute of Microelectronics, Chinese Academy of Sciences, Beijing, 100029…. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Key Technologies Research and Development Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Performing visual recognition and combinatorial optimization simultaneously on a single multifunctional neuromorphic computing platform offers significant advantages in terms of efficiency, real‐time processing, and integrated decision‐making. However, the advances are hindered by hardware constraints. Here, a new type of all‐electrically controlled labyrinth magnetic texture (MT) devices is reported, wherein the trainable MT can be created, manipulated, and detected efficiently and reliably at room temperature. By utilizing the spin‐orbit torque (SOT) effect, it can modulate the nonlinear magneto‐resistance of the device via a corroborated dynamic conductance matrix, mimicking the mixed short‐term and long‐term potentiation of biological synapses. The developed SOT‐MT devices excel in diverse in‐memory computing tasks, including pattern recognition and combinatorial optimization. Utilizing a cross‐bar array with single SOT‐MT devices, outstanding test accuracy is achieved over 93% on MNIST, and a success rate exceeding 95% in solving the 8‐city traveling salesman problem with the Hopfield network. Synergistic tailoring of constant and dynamic fluctuations contributes to this success. The study paves the way for dynamic network MT devices, advancing complex task processing by enabling efficient fusion of cognition and combinatorial optimization on a single neuromorphic hardware system.

Schlagwörter

Stichwörter der Autor:innen

advanced neuromorphic hardwarecombinatorial optimization problemsdomain wall motionmagnetic texturepattern recognitionspin‐orbit torque

Autor:innen und Institutionen

  • State Key Laboratory of Fabrication Technologies for Integrated Circuits, Institute of Microelectronics, Chinese Academy of Sciences, Beijing, 100029, China.
  • University of Chinese Academy of Sciences, Beijing, 100049, China.
  • VeriSilicon Technology (Shanghai) Co., Ltd., Shanghai, 201306, China.
  • School of Civil Engineering and Architecture, Northeast Electric Power University, Jilin, 132012, China.
  • Jilin Special Equipment Inspection and Research Institute, Jilin, 132013, China.
  • Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang, 110016, China.

ORCID-Profile: Guozhong Xing

Förderung

  • National Natural Science Foundation of China
  • Key Technologies Research and Development Program

Zitieren

APA

Zhang, Y., Li, Y., Lin, H., Wang, X., Liu, L., Xing, G., Wang, D., Zhao, Z., Guo, Z., Yang, J., Niu, J., Sun, Y., Wu, T., & Xing, G. (2026). Harnessing Time‐Dependent Magnetic Texture Dynamics via Spin‐Orbit Torque for Physics‐Enhanced Neuromorphic Computing. Advanced Science, 13(12). https://doi.org/10.1002/advs.202513946

Vancouver

Zhang Y, Li Y, Lin H, Wang X, Liu L, Xing G, et al. Harnessing Time‐Dependent Magnetic Texture Dynamics via Spin‐Orbit Torque for Physics‐Enhanced Neuromorphic Computing. Advanced Science. 2026;13(12). doi:10.1002/advs.202513946

BibTeX
@article{Zhang2026,
  author = {Zhang, Yifan and Li, Yu and Lin, Huai and Wang, Xinying and Liu, Long and Xing, Guoliang and Wang, Di and Zhao, Zhihao and Guo, Zhipeng and Yang, Junbo and Niu, Jiebin and Sun, Yan and Wu, Tom and Xing, Guozhong},
  title = {Harnessing Time‐Dependent Magnetic Texture Dynamics via Spin‐Orbit Torque for Physics‐Enhanced Neuromorphic Computing},
  journal = {Advanced Science},
  year = {2026},
  volume = {13},
  number = {12},
  doi = {10.1002/advs.202513946}
}
Kostenlos starten →