Easy-plane spin Hall nano-oscillators as spiking neurons for neuromorphic computing

Published

January 10, 2022

Author(s)

Danijela Marković, Matthew Daniels, Pankaj Sethi, Andrew Kent, Mark Stiles, Julie Grollier

Abstract

We show analytically using a macrospin approximation that easy-plane spin Hall nano-oscillators excited by a spin-current polarized perpendicularly to the easy-plane have phase dynamics analogous to that of Josephson junctions. This allows them to reproduce the spiking behavior of biological neurons that is highly demanded for neuromorphic computing. We perform micromagnetic simulations of such oscillators realized in the nanoconstriction geometry and show that the Josephson dynamics is preserved in an experimentally feasible architecture. Finally, we show that such spiking spintronic neurons can be coupled in chain networks through synapses whose weights can be controlled by locally modifying the anisotropy, and thus used for neuromorphic computing.

Citation

Physical Review B

Volume

105

Issue

1

Pub Type

Journals

Download Paper

https://doi.org/10.1103/PhysRevB.105.014411

Local Download

Citation

Markovic, D. , Daniels, M. , Sethi, P. , Kent, A. , Stiles, M. and Grollier, J. (2022), Easy-plane spin Hall nano-oscillators as spiking neurons for neuromorphic computing, Physical Review B, [online], https://doi.org/10.1103/PhysRevB.105.014411, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=933138 (Accessed October 15, 2024)