Reservoir Computing with Frustrated Nanomagnet Arrays

ORAL

Abstract

Reservoir computing (RC) [1] is a subset of recurrent neural network where only the weights of the output layer are updated during training. This technique is therefore well suited for resource constrained hardware environments. We propose a novel reservoir comprising a planar arrangement of nanomagnets each having perpendicular magnetic anisotropy (PMA) [2]. The effect of nanomagnet magnetic fields upon adjacent nanomagnets exhibits two features: non-linear interaction and variable interaction strength, making the proposed implementation well suited for RC. Information is input by stimulating individual nanomagnets with spin-torque. The magnetizations of various nanomagnets are read electrically via magnetic tunnel junctions. A trained single layer circuit is used to perform vector-matrix multiplication on the magnetization values and the output weights to obtain the output vector. The nanomagnet reservoir was simulated in mumax3 with an input stream comprising triangle or square waves. The reservoir successfully identified the waveforms with 100% accuracy for both the training and testing data.

[1] H. Jaeger, Bonn, Germany: Germ. Nat. Res. Cent. Info. Tech. GMD Tech. Rep., Vol. 148, p. 13 (2001)
[2] P. Zhou et al. arXiv:cs.NE/2003.10948

*With funding from the AFRL, NSF CCF #1815033

–

Presenters

  • Alexander Edwards

    • Department of Electrical and Computer Engineering, University of Texas at Dallas

Authors

  • Alexander Edwards

    • Department of Electrical and Computer Engineering, University of Texas at Dallas

  • Peng Zhou

    • Department of Electrical and Computer Engineering, University of Texas at Dallas

  • Dhritiman Bhattacharya

    • Virginia Commonwealth Univ
    • Department of Mechanical and Nuclear Engineering, Virginia Commonwealth University

  • Nathan R. McDonald

    • Information Directorate, Air Force Research Laboratory

  • Felipe Garcia-Sanchez

    • Universidad de Salamanca
    • University of Salamanca
    • Applied Physics, Universidad de Salamanca
    • Department of Applied Physics, Universidad de Salamanca

  • Lisa Loomis

    • Information Directorate, Air Force Research Laboratory

  • Clare D. Thiem

    • Information Directorate, Air Force Research Laboratory

  • Jayasimha Atulasimha

    • Virginia Commonwealth Univ
    • Department of Mechanical and Nuclear Engineering, Virginia Commonwealth University

  • Joseph S. Friedman

    • Electrical and Computer Engineering, University of Texas at Dallas
    • University of Texas at Dallas
    • Electrical and Computer Engineering Dept., University of Texas at Dallas, Richardson TX USA
    • Electrical & Computer Engineering, University of Texas at Dallas
    • Department of Electrical and Computer Engineering, University of Texas at Dallas