Presentation Information
[22p-A302-6]Development of magneto-optical recording techinique for magneto-optical diffractive deep neural networks
〇(DC)Hotaka Sakaguchi1, Jian Zhang1, Riku Oya1, Reo Akagawa1, Juri Ikeda1, Satoshi Sumi2, Hiroyuki Awano2, Hirofumi Nonaka3, Fatima Zahra Chafi1, Takayuki Ishibashi1 (1.Nagaoka Univ. of Tech., 2.Toyota Tech. Inst., 3.Aichi Inst. of Tech.)
Keywords:
optical neural network,magneto-optical effect,magnetic thin film
Recently, reservoir computing and neural networks that utilise physical phenomena have attracted attention for their low power consumption and high speed computation in machine learning. We proposed a magneto-optical diffraction deep neural networks that utilise the diffraction of light and the magneto-optical effect of magnetic materials, and reported on its visible light operation and its fabrication technique. The connectivity of neurons between layers and the density of neurons can be increased with smaller magnetic domain size, which is expected to improve performance. In this study, magnetic domain recording conditions were investigated for the formation of small magnetic domains.