Regensburg 2025 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MA: Fachverband Magnetismus
MA 41: Poster III
MA 41.3: Poster
Thursday, March 20, 2025, 15:00–17:30, P3
Current Induced Markov State Modeling with Skyrmions in Geometrical Confinements — Thomas Brian Winkler1, Grischa Beneke2, •Yuean Zhou2, Johan Mentink1, and Mathias Kläui2 — 1Radboud University, Institute for Molecules and Materials, Netherlands — 2Institut für Physik, Johannes Gutenberg-Universität Mainz, Germany
Magnetic skyrmion in geometrical confinements have recently gained attention as a platform for neuromorphic computing for instance for Brownian Reservoir Computing [1, 2]. Despite significant progress with coarse-grained particle-based modeling of the stochastic dynamics of skyrmion [3], the interplay between stochasticity and device defects remains underexplored. Here, we extract the stochastic transitions of skyrmion between pinning sites using a simple Markov state framework. The resulting Markov state model enables system behavior prediction based on the effective energy landscape of the device. The time step for Markov state evolution is significantly faster than traditional simulations, which makes the system well-suited for leveraging the digital twin for predicting experimental outcomes and evaluating reservoir metrics. We also aim to implement energy-efficient Markov chain Monte Carlo algorithms in-materio.
[1] K. Raab et al., Nat. Commun. 13 (1), 6982 (2022)
[2] G. Beneke et al., Nat. Commun. 15 (1), 8103 (2024)
[3] T. B. Winkler et al., Appl. Phys. Lett. 124 (2), 022403 (2024)
Keywords: Skyrmion; Markov State Model