A non-overlapped implantation MOSFET differential pair implementation of bidirectional weight update synapse for neuromorphic computing

被引：5

作者：

Jeng, E. S. ^{[1
]}

Chen, H. X. ^{[1
]}

Chiang, Y. L. ^{[1
]}

Chang, J. H. ^{[1
]}

Chen, J. Y. ^{[1
]}

机构：

[1] Chung Yuan Christian Univ, Dept Elect Engn, Chungli 32023, Taiwan

来源：

MICROELECTRONICS JOURNAL | 2019年 / 90卷

关键词：

Neuromorphics; Neural network hardware; Analog memory; Artificial synapse; SINGLE; NMOSFETS; SPACERS; CHIP;

D O I：

10.1016/j.mejo.2019.07.001

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

A non-volatile memory (NVM) differential pair realization of the synaptic weight for an artificial neural network (ANN) circuit is investigated. Two non-overlapped implantation (NOI) nMOSFETs are proposed to form a NVM differential pair as an artificial synapse. The pair of NOI transistors are non-volatile analog memories and capable of storing positive or negative synaptic weight. In this study, a NOI differential pair based silicon neural chip of three neurons with 36 analog synapses in total is designed, simulated, fabricated, and verified by the chip-in-the-loop learning process. The classification performance of the neural chip when training and testing with the IRIS dataset is reported. This differential pair design not only overcomes the potential constraint in the weights of NOI but also exhibits better classification performances than that of single NOI-based ANN.

引用

页码：306 / 314

页数：9

共 23 条

[1]

Aamir S. A., 2018, IEEE T CIRCUITS-I, P1

[2] High precision tuning of state for memristive devices by adaptable variation-tolerant algorithm [J].

Alibart, Fabien ;

Gao, Ligang ;

Hoskins, Brian D. ;

Strukov, Dmitri B. .

NANOTECHNOLOGY, 2012, 23 (07)

[3] Unsupervised Learning by Spike Timing Dependent Plasticity in Phase Change Memory (PCM) Synapses [J].

Ambrogio, Stefano ;

Ciocchini, Nicola ;

Laudato, Mario ;

Milo, Valerio ;

Pirovano, Agostino ;

Fantini, Paolo ;

Ielmini, Daniele .

FRONTIERS IN NEUROSCIENCE, 2016, 10

[4] A single-transistor silicon synapse [J].

Diorio, C ;

Hasler, P ;

Minch, A ;

Mead, CA .

IEEE TRANSACTIONS ON ELECTRON DEVICES, 1996, 43 (11) :1972-1980

[5] Computational intelligence methods for rule-based data understanding [J].

Duch, W ;

Setiono, R ;

Zurada, JM .

PROCEEDINGS OF THE IEEE, 2004, 92 (05) :771-805

[6]

Garbin D, 2015, IEEE INT SYMP NANO, P193, DOI 10.1109/NANOARCH.2015.7180611

[7] NVM characteristics of single-MOSFET cells using nitride spacers with gate-to-drain NOI [J].

Hsieh, CS ;

Kao, PC ;

Chin, CS ;

Hon, CH ;

Fan, CC ;

Kung, WC ;

Wang, ZW ;

Jeng, ES .

IEEE TRANSACTIONS ON ELECTRON DEVICES, 2004, 51 (11) :1811-1817

[8] Chip implementation of supervised neural network using single-transistor synapses [J].

Jeng, E. S. ;

Chou, S. W. ;

Chen, H. X. ;

Chiang, Y. L. .

MICROELECTRONICS JOURNAL, 2017, 66 :76-83

[9] Non-overlapped implantation (NOI) MOSFET synapse and its implementation on supervised neural network [J].

Jeng, E. S. ;

Peng, K. M. ;

Chou, S. W. ;

Chen, H. X. ;

Chiang, Y. L. .

NEUROCOMPUTING, 2015, 167 :290-298

[10] Characterization of single-sided gate-to-drain non-overlapped implantation nMOSFETs for multi-functional non-volatile memory applications [J].

Jeng, E. S. ;

Chen, Y. F. ;

Chang, C. C. ;

Peng, K. M. ;

Chou, S. W. ;

Ho, C. W. ;

Huang, C. F. ;

Gong, J. .

SOLID-STATE ELECTRONICS, 2012, 68 :73-79

← 1 2 3 →