3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation

I. Ting Wang; Yen Chuan Lin; Yu Fen Wang; Chung Wei Hsu; Tuo-Hung Hou

doi:10.1109/IEDM.2014.7047127

3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation

I. Ting Wang
, Yen Chuan Lin
, Yu Fen Wang
, Chung Wei Hsu
, Tuo-Hung Hou

Research output: Contribution to journal › Conference article › peer-review

48 Scopus citations

Abstract

A high-density 3D synaptic architecture based on self-rectifying Ta/TaO_x/TiO₂/Ti RRAM is proposed as an energy- and cost-efficient neuromorphic computation hardware. The device shows excellent analog synaptic features that can be accurately described by the physical and compact models. Ultra-low energy consumption comparable to that of a biological synapse (<10 fJ/spike) has been demonstrated for the first time.

Original language	English
Article number	7047127
Pages (from-to)	28.5.1-28.5.4
Number of pages	4
Journal	Technical Digest - International Electron Devices Meeting, IEDM
Volume	2015-February
Issue number	February
DOIs	https://doi.org/10.1109/IEDM.2014.7047127
State	Published - 15 Dec 2015
Event	2014 60th IEEE International Electron Devices Meeting, IEDM 2014 - San Francisco, United States Duration: 15 Dec 2014 → 17 Dec 2014

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Access to Document

10.1109/IEDM.2014.7047127

Cite this

@article{73bd632db5224e4ea8bde2b2d0fa3f39,

title = "3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation",

abstract = "A high-density 3D synaptic architecture based on self-rectifying Ta/TaOx/TiO2/Ti RRAM is proposed as an energy- and cost-efficient neuromorphic computation hardware. The device shows excellent analog synaptic features that can be accurately described by the physical and compact models. Ultra-low energy consumption comparable to that of a biological synapse (<10 fJ/spike) has been demonstrated for the first time.",

author = "Wang, \{I. Ting\} and Lin, \{Yen Chuan\} and Wang, \{Yu Fen\} and Hsu, \{Chung Wei\} and Tuo-Hung Hou",

year = "2015",

month = dec,

day = "15",

doi = "10.1109/IEDM.2014.7047127",

language = "English",

volume = "2015-February",

pages = "28.5.1--28.5.4",

journal = "Technical Digest - International Electron Devices Meeting, IEDM",

issn = "0163-1918",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "February",

note = "2014 60th IEEE International Electron Devices Meeting, IEDM 2014 ; Conference date: 15-12-2014 Through 17-12-2014",

}

TY - JOUR

T1 - 3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation

AU - Wang, I. Ting

AU - Lin, Yen Chuan

AU - Wang, Yu Fen

AU - Hsu, Chung Wei

AU - Hou, Tuo-Hung

PY - 2015/12/15

Y1 - 2015/12/15

N2 - A high-density 3D synaptic architecture based on self-rectifying Ta/TaOx/TiO2/Ti RRAM is proposed as an energy- and cost-efficient neuromorphic computation hardware. The device shows excellent analog synaptic features that can be accurately described by the physical and compact models. Ultra-low energy consumption comparable to that of a biological synapse (<10 fJ/spike) has been demonstrated for the first time.

AB - A high-density 3D synaptic architecture based on self-rectifying Ta/TaOx/TiO2/Ti RRAM is proposed as an energy- and cost-efficient neuromorphic computation hardware. The device shows excellent analog synaptic features that can be accurately described by the physical and compact models. Ultra-low energy consumption comparable to that of a biological synapse (<10 fJ/spike) has been demonstrated for the first time.

UR - https://www.scopus.com/pages/publications/84938273766

U2 - 10.1109/IEDM.2014.7047127

DO - 10.1109/IEDM.2014.7047127

M3 - Conference article

AN - SCOPUS:84938273766

SN - 0163-1918

VL - 2015-February

SP - 28.5.1-28.5.4

JO - Technical Digest - International Electron Devices Meeting, IEDM

JF - Technical Digest - International Electron Devices Meeting, IEDM

IS - February

M1 - 7047127

T2 - 2014 60th IEEE International Electron Devices Meeting, IEDM 2014

Y2 - 15 December 2014 through 17 December 2014

ER -

3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this