Skip to main content
  • Paul Scherrer Institut PSI
  • PSI Research, Labs & User Services

Digital User Office

  • Digital User Office
  • DE
  • EN
  • FR
Paul Scherrer Institut (PSI)
Search
Paul Scherrer Institut (PSI)

Hauptnavigation

  • Research at PSIOpen mainmenu item
    • Research Initiatives
    • Ethics and Research integrity
    • Scientific Highlights
    • Scientific Events
    • Scientific Career
    • PSI-FELLOW
    • PSI Data Policy
  • Research Divisions and LabsOpen mainmenu item
    • Overview
    • Research with Neutrons and Muons
    • Photon Science
    • Energy and Environment
    • Nuclear Energy and Safety
    • Biology and Chemistry
    • Scientific Computing, Theory and Data
    • Large Research Facilities
  • Facilities and InstrumentsOpen mainmenu item
    • Overview
    • Large Research Facilities
    • Facilities
    • PSI Facility Newsletter
  • PSI User ServicesOpen mainmenu item
    • User Office
    • Methods at the PSI User Facilities
    • Proposals for beam time
    • Proposal Deadlines
    • Data Analysis Service (PSD)
    • EU support programmes
  • New ProjectsOpen mainmenu item
    • SLS 2.0
    • IMPACT
  • DE
  • EN
  • FR

Digital User Office (mobile)

  • Digital User Office

You are here:

  1. PSI Home
  2. Labs & User Services
  3. PSD
  4. LSC
  5. MicroSpec Group
  6. Scientific Highlights and News
  7. Can skyrmions read?

Secondary navigation

MicroSpec

  • People
  • Research
  • Scientific Highlights and News
  • Pixelator
  • Publications
17 March 2020

Can skyrmions read?

Sketch of the operations principle of a skyrmion artificial neural network
Sketch of the operation principle of a skyrmionic artificial neural network. (a) Sketch of the operation concept, where the presence or absence of magnetic skyrmions in a reservoir is used to modulate the weight of a synapse in an artificial neural network. (b) Sketch of a simulated skyrmionic artificial neural network employed to "read" a handwritten number.

Magnetic skyrmions are spin configurations that have become extremely popular with the spintronics research community in the last few years. Their popularity is due to a set of interesting properties that these magnetic states exhibit, which originate out of their non-trivial topology. A non-exhaustive list of these properties include efficient current-induced manipulation, a resistance against annihilation from pinning sites and defects, and the possibility to nucleate these spin states with diameters down to a few tens of nm. These properties have led to numerous proposals to use magnetic skyrmions as encoders for digital information, where the digital bits 1/0 are encoded in the presence/absence of the magnetic skyrmion.

Another potential application for magnetic skyrmions, exploiting some of their more exotic topology-induced properties, is to employ them to generate synaptic weights in artificial neural networks. This proposal was tested by an international collaboration of scientists, led by the Korea Institute of Technology and by the IBM Watson research center, who injected magnetic skyrmions into a patterned reservoir, and used the changes in the anomalous Hall effect caused by a change in the number of skyrmions present in the reservoir to change the synaptic weight in a one-neuron proof-of-principle neural network.

To visualize the magnetic configuration in the skyrmion reservoir, and to be able to directly correlate the change in the number of skyrmions present in the reservoir to the change in the anomalous Hall resistance, the researchers utilized the scanning X-ray transmission microscope installed at the PolLux endstation of the Swiss Light Source. Thanks to the combined magnetic imaging and electrical transport setup developed at the PolLux endstation, the researchers could observe that the synaptic weight of the patterned reservoir can be reliably and reproducibly changed by modifying the number of skyrmions present.

Using the data measured at the PolLux endstation, the research team performed device-level and system-level simulations of a skyrmionic artificial neural network, which was trained on pattern recognition using the MNIST handwritten dataset. The simulations resulted, after training, in a pattern recognition accuracy of 89%, comparable to the ideal software-based accuracy of 93%. Furthermore, system-level simulations revealed that a skyrmionic neural network could have important advantages compared to other technologies, such as e.g. an improved endurance and retention.

With these results, a first proof-of-principle of the applicability of magnetic skyrmions for neuromorphic computing was demonstrated.

Contacts:
Dr. Jörg Raabe
Swiss Light Source
Paul Scherrer Institut
Telephone: +41 56 310 5193
E-mail: joerg.raabe@psi.ch

Dr. Seonghoon Woo
IBM T.J. Watson Research Center
Telephone: +1 914-945-2572
E-mail: shwoo@ibm.com

Original Publication:
Skyrmion-based artificial synapses for neuromorphic computing
K.-M. Song, J.-S. Jeong, B. Pan, X. Zhang, J. Xia, S. Cha, T.-E. Park, K. Kim, S. Finizio, J. Raabe, J. Chang, Y. Zhou, W. Zhao, W. Kang, H. Ju, and S. Woo
Nature Electronics (2020)
DOI:
10.1038/s41928-020-0385-0

Other Media:
Highlight on the Device and Materials Engineering community – Nature Publishing Group
https://devicematerialscommunity.nature.com/users/357990-kyung-mee-song/posts/61801-magnetic-skyrmions-extend-their-potential-to-neuromorphic-computing-for-future-ai

Sidebar

top

Footer

Paul Scherrer Institut

Forschungsstrasse 111
5232 Villigen PSI
Switzerland

Telephone: +41 56 310 21 11
Telefax: +41 56 310 21 99

How to find us
Contact

Visitor Centre psi forum
School Lab iLab (in German)
Center for Proton Therapy
PSI Education Centre
PSI Guest House
PSI Gastronomie (in German)
psi forum shop

Service & Support

  • Phone Book
  • User Office
  • Accelerator Status
  • PSI Publications
  • Suppliers
  • E-Billing
  • Computing
  • Safety (in German)

Career

  • Working at PSI
  • Job Opportunities
  • Training and further education
  • Career Center
  • Vocational Training (in German)
  • PSI Education Center

For the media

  • PSI in brief
  • Facts and Figures
  • Media corner
  • Media Releases
  • Social Media

Follow us: Twitter (in English) LinkedIn Youtube Facebook Instagram Issuu RSS

Footer legal

  • Imprint
  • Terms and Conditions
  • Editors' login