Berlin 2024 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz

AKPIK 1: Reservoir Computing & Neural Networks

AKPIK 1.2: Vortrag

Dienstag, 19. März 2024, 09:45–10:00, MAR 0.002

Novel implementations for reservoir computing -- from spin to charge — •Atreya Majumdar¹, Karin Everschor-Stte¹, Katharina Wolk², and Dennis Meier^{2, 3} — ¹Faculty of Physics and Center for Nanointegration Duisburg-Essen (CENIDE), University of Duisburg-Essen, 47057 Duisburg, Germany — ²Department of Materials Science and Engineering, Norwegian University of Science and Technology (NTNU), Trondheim 7034, Norway — ³Center for Quantum Spintronics, Norwegian University of Science and Technology (NTNU), Trondheim 7034, Norway

Magnetic and ferroelectric materials are emerging as promising candidates for unconventional computing and next-generation information technology. We review and explore the potential of nanoscale topological textures, focusing on magnetic skyrmions and ferroelectric domain walls, for use in reservoir computing [1] a scheme that allows transforming non-linear tasks into linearly solvable ones. We highlight the essential characteristics needed for physical reservoirs, outlining the advantages of topological textures, such as the increased complexity and flexible input and output options. We provide insights into how topological textures in magnetic and ferroelectric systems can serve as an avenue for enhancing reservoir computing and, more generally, broadening the scope of in-materio computing.

[1] K. Everschor-Sitte, A. Majumdar, K. Wolk, D. Meier, arXiv:2311.11929

Keywords: Physical reservoir computing; Skyrmions; Ferroelectric materials; Neuromorphic computing; Topological textures