Memristor crossbar array for binarized neural networks

Yong Kim; Won Hee Jeong; Son Bao Tran; Hyo Cheon Woo; Jihun Kim; Cheol Seong Hwang; Kyeong Sik Min; Byung Joon Choi

doi:10.1063/1.5092177

Memristor crossbar array for binarized neural networks

Yong Kim
, Won Hee Jeong
, Son Bao Tran
, Hyo Cheon Woo
, Jihun Kim
, Cheol Seong Hwang
, Kyeong Sik Min
, Byung Joon Choi

Dept. of Materials Science & Engineering

Research output: Contribution to journal › Article › peer-review

29 Scopus citations

Abstract

Memristor crossbar arrays were fabricated based on a Ti/HfO₂/Ti stack that exhibited electroforming-free behavior and low device variability in a 10 x 10 array size. The binary states of high-resistance-state and low-resistance-state in the bipolar memristor device were used for the synaptic weight representation of a binarized neural network. The electroforming-free memristor was confirmed as being suitable as a binary synaptic device because of its higher device yield, lower variability, and less severe malfunction (for example, hard break-down) than the electroformed memristors based on a Ti/HfO₂/Pt structure. The feasibly working binarized neural network adopting the electroforming-free binary memristors was demonstrated through simulation.

Original language	English
Article number	045131
Journal	AIP Advances
Volume	9
Issue number	4
DOIs	https://doi.org/10.1063/1.5092177
State	Published - 1 Apr 2019

Access to Document

10.1063/1.5092177

Cite this

@article{e4e7461a1f83468580b419ef04b36a60,

title = "Memristor crossbar array for binarized neural networks",

abstract = "Memristor crossbar arrays were fabricated based on a Ti/HfO2/Ti stack that exhibited electroforming-free behavior and low device variability in a 10 x 10 array size. The binary states of high-resistance-state and low-resistance-state in the bipolar memristor device were used for the synaptic weight representation of a binarized neural network. The electroforming-free memristor was confirmed as being suitable as a binary synaptic device because of its higher device yield, lower variability, and less severe malfunction (for example, hard break-down) than the electroformed memristors based on a Ti/HfO2/Pt structure. The feasibly working binarized neural network adopting the electroforming-free binary memristors was demonstrated through simulation.",

author = "Yong Kim and Jeong, \{Won Hee\} and Tran, \{Son Bao\} and Woo, \{Hyo Cheon\} and Jihun Kim and Hwang, \{Cheol Seong\} and Min, \{Kyeong Sik\} and Choi, \{Byung Joon\}",

note = "Publisher Copyright: {\textcopyright} 2019 Author(s).",

year = "2019",

month = apr,

day = "1",

doi = "10.1063/1.5092177",

language = "English",

volume = "9",

journal = "AIP Advances",

issn = "2158-3226",

number = "4",

}

TY - JOUR

T1 - Memristor crossbar array for binarized neural networks

AU - Kim, Yong

AU - Jeong, Won Hee

AU - Tran, Son Bao

AU - Woo, Hyo Cheon

AU - Kim, Jihun

AU - Hwang, Cheol Seong

AU - Min, Kyeong Sik

AU - Choi, Byung Joon

PY - 2019/4/1

Y1 - 2019/4/1

N2 - Memristor crossbar arrays were fabricated based on a Ti/HfO2/Ti stack that exhibited electroforming-free behavior and low device variability in a 10 x 10 array size. The binary states of high-resistance-state and low-resistance-state in the bipolar memristor device were used for the synaptic weight representation of a binarized neural network. The electroforming-free memristor was confirmed as being suitable as a binary synaptic device because of its higher device yield, lower variability, and less severe malfunction (for example, hard break-down) than the electroformed memristors based on a Ti/HfO2/Pt structure. The feasibly working binarized neural network adopting the electroforming-free binary memristors was demonstrated through simulation.

AB - Memristor crossbar arrays were fabricated based on a Ti/HfO2/Ti stack that exhibited electroforming-free behavior and low device variability in a 10 x 10 array size. The binary states of high-resistance-state and low-resistance-state in the bipolar memristor device were used for the synaptic weight representation of a binarized neural network. The electroforming-free memristor was confirmed as being suitable as a binary synaptic device because of its higher device yield, lower variability, and less severe malfunction (for example, hard break-down) than the electroformed memristors based on a Ti/HfO2/Pt structure. The feasibly working binarized neural network adopting the electroforming-free binary memristors was demonstrated through simulation.

UR - https://www.scopus.com/pages/publications/85065211843

U2 - 10.1063/1.5092177

DO - 10.1063/1.5092177

M3 - Article

AN - SCOPUS:85065211843

SN - 2158-3226

VL - 9

JO - AIP Advances

JF - AIP Advances

IS - 4

M1 - 045131

ER -

Memristor crossbar array for binarized neural networks

Abstract

Access to Document

Other files and links

Fingerprint

Cite this