Presentation Information
[25a-P01-26]Investigating the cluster spin glass behavior in garnet based ferromagnet towards spin waves based neuromorphic computation
〇(PC)shamim sarker1, Haining Li1, EMK Ikball Ahamed1, Hiroyasu Yamahara1, Siyi Tang1, Zhiqiang Liao1, Tetsuya Iizuka1, Munetoshi Seki1, Hitoshi Tabata1 (1.Tokyo Univ)
Keywords:
Spin wave,Spin glass,Magnetics
Spin glass (SG) exhibits a frozen spin state at low temperatures due to competing random magnetic interactions and shows photoinduced magnetism and an aging memory effect. Random magnetic interaction and associated aging memory render history-dependent properties that can offer the opportunity to mimic the human brain's functionality (based on a similar Hamiltonian). On the other hand, the spin waves (SWs) offer extremely low power computation due to the absence of translational motion. Our study aims to combine the low power property of SWs and the memory property of SG to implement ultra-low power neuromorphic computation (i.e., physical reservoir computation). However, spin frozen state in SG appears only at extremely low temperatures, inhibiting its practical and low power application. Therefore, the primary focus of this research is to investigate the possibility of achieving optimum room temperature spin glass material extracted from the order controlling of garnet-based ferromagnetic materials. From consideration of magnetization dynamics in further study, we started with yttrium iron garnet (YIG) due to its lowest reported Gilbert damping constant. For the initial investigation, we have prepared two YIG thin films with thicknesses of 16 and 40 nm by PLD.