2017年第78回応用物理学会秋季学術講演会

講演情報

一般セッション(口頭講演)

9 応用物性 » 9.2 ナノワイヤ・ナノ粒子

[6p-S44-1~21] 9.2 ナノワイヤ・ナノ粒子

2017年9月6日(水) 13:15 〜 19:00 S44 (第5会議室)

深田 直樹(物材機構)、石川 史太郎(愛媛大)、清水 智弘(関西大)

17:15 〜 17:30

[6p-S44-15] Novel application of memristive nanowire complex network for brain-like computation

〇(D)QIAO LI1,2、Rintaro Higuchi2、Yoshitaka Shigaya2、Tomonobu Nakayama1,2 (1.Univ. of Tsukuba、2.MANA,NIMS)

キーワード:nanowire, brain-like computation

Neuromorphic network based on the artificial synaptic network is expected to memory and analyze information with high performance as human brain works. Resistance memristor which shows short-term plasticity and long-term potentiation is a promising candidate as artificial synapse and has been widely studied recently. The main problem to achieve brain-like computation is how to train solid state materials to learn. In this work, we propose to use the memristive-based complex network for future brain-like computation. NiO/Ni nanowire is chosen as it is a well-known memristive resistor and can be easily synthesized. From electrical measurement, NiO/Ni nanowire complex network shows clear memristive behavior and can work as multi-state resistance switch. Training on the network is still on the go.