Regensburg 2022 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MA: Fachverband Magnetismus
MA 19: Poster 1
MA 19.30: Poster
Tuesday, September 6, 2022, 17:30–20:00, P2
Brownian reservoir computing realized using geometrically confined skyrmions — •Klaus Raab1, Maarten A. Brems1, Grischa Beneke1, Takaaki Dohi1, Jan Rothörl1, Fabian Kammerbauer1, Johan H. Mentink2, and Mathias Kläui1,3 — 1Institut für Physik, Johannes Gutenberg-Universität Mainz, Staudingerweg 7, 55128 Mainz, Germany — 2Radboud University, Institute for Molecules and Materials, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands — 3Graduate School of Excellence Materials Science in Mainz, Staudingerweg 9, 55128 Mainz, Germany
We demonstrate experimentally [1] a conceptionally new approach for reservoir computing (RC), that leverages the thermally activated diffusive motion of magnetic skyrmions in a confined, triangular geometry. The combination of gated and thermal skyrmion motion of a single skyrmion already suffices to realize all Boolean logic gate operations including the non-linearly separable XOR operation that cannot be realized using a conventional single layer perceptron. An effective potential well created by the confinement allows for a natural, energy efficient reset mechanism enabled by the thermal fluctuations of the skyrmions. We show that the output training costs using linear regression are low and that our ultra-low power operation using current densities orders of magnitude smaller than used in existing spintronic reservoir computing demonstrations. Our concept can be easily extended by linking multiple confined geometries for scalable and low-energy reservoir computing. [1] K. Raab et al. Brownian reservoir computing with geometrically confined skyrmions - arXiv:2203.14720