[16p-301-1~10] The role of functional oxides in the next-generation neuromorphic hardware

[16p-301-5] Challenges in Silicon Neuronal Networks

〇Takashi Kohno¹ (1.IIS, Univ. Tokyo)

Keywords:neuromorphic system, artificial intelligence

Silicon neuronal networks reproduce the electrophysiological activities in the nervous system by the networks of the electronic circuits that correspond to neuronal cells and synapses. Their internal information expression is expected to have a high degree of compatibility with the nervous system, which allows them to realize not only a neuromimetic adaptive and robust information processing with ultra-low power consumption but also a seamless information sharing and communication with the nervous system. They are a candidate for the core technology in the next generation artificial intelligence that can also be applied to neurological devices. Current status and challenges in this field will be explained.