The 80th JSAP Autumn Meeting 2019

Presentation information

Oral presentation

Code-sharing session » 【CS.8】 Code-sharing Session of 10.1, 10.2, 10.3 & 10.4

[20p-E216-1~7] 【CS.8】 Code-sharing Session of 10.1, 10.2, 10.3 & 10.4

Fri. Sep 20, 2019 1:30 PM - 3:15 PM E216 (E216)

Eiji Shikoh(大阪市大)

2:15 PM - 2:30 PM

[20p-E216-4] Antiferromagnet/Ferromagnet Heterostructures for Artificial Neurons and Synapses

Aleksandr Kurenkov1,2,3, Samik DuttaGupta1,2,3, Chaoliang Zhang1,4,5, Shunsuke Fukami1,2,3,5,6, Yoshihiko Horio1, Hideo Ohno1,2,3,5,6 (1.RIEC, Tohoku University, 2.CSIS, 3.CSRN, 4.FRIS, 5.CIES, 6.WPI-AIMR)

Keywords:neuromorphic computing, spin-orbit torque switching

Time-governed properties of synapses and neurons are the foundation of information processing in the human brain.[1,2] Thus emulation of such properties in artificial devices is the key to building compact and efficient artificial synapses and neurons required for future large-scale artificial neural networks. Here we show how it can be achieved in antiferromagnet/ferromagnet PtMn/[Co/Ni] heterostructures[3] by using the dynamics of spin-orbit torque switching. For that we investigate the switching by pulses of widths from 1 s to 1 ns and demonstrate that it naturally allows measuring pulse-to-pulse time intervals. Based on this, we realize biologically plausible neuron- and synapse-like operation in simple devices.[4] Their operation, in contrast to other works, is inherent and can be driven by pulses of virtually any shape. The efficiency has the potential to reach that of the human brain. Furthermore, the previously shown size-dependent switching of the material system[5] allows making both the synapse- and the neuron-like elements from the same material stack. This uniformity in fabrication and coherency of operation is crucial for forming a hardware basis for future large-scale artificial spiking neural networks.
A portion of this work was supported by JSPS KAKENHI 17H06093, 18KK0143 and 19K15428, JST-OPERA, JSPS Core-to-Core Program and Cooperative Research Projects of RIEC.
[1] N. K. Kasabov, Neural Networks 52, 62 (2014).
[2] S. Song, K. D. Miller, L. F. Abbott, Nat. Neurosci. 3, 919 (2000).
[3] S. Fukami, C. Zhang, S. DuttaGupta, A. Kurenkov, H. Ohno, Nat. Mater. 15, 535 (2016).
[4] A. Kurenkov, S. DuttaGupta, C. Zhang, S. Fukami, Y. Horio, H. Ohno, Adv. Mater. 31, 1900636
(2019).
[5] A. Kurenkov, C. Zhang, S. DuttaGupta, S. Fukami, H. Ohno, Appl. Phys. Lett. 110, 092410 (2017).