Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network

Seung Cheol Han; Jonghyun Choi; Sung Min Hong

doi:10.1109/SISPAD54002.2021.9592540

Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network

Seung Cheol Han, Jonghyun Choi, Sung Min Hong

Department of Artificial Intelligence

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We propose to use a 3D convolutional neural network to accelerate three-dimensional device simulation by generating an electrostatic potential profile. In the training phase, the deep neural network is trained with the simulation results for various 3D MOSFETs in a supervised manner. The generated potential profile is used as an initial guess at a non-equilibrium condition, while carrier densities are estimated by the frozen field simulation. By numerical examples for three-dimensional MOSFETs, we show that the proposed method significantly reduces the number of the Newton iterations.

Original language	English
Title of host publication	SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	52-55
Number of pages	4
ISBN (Electronic)	9781665406857
DOIs	https://doi.org/10.1109/SISPAD54002.2021.9592540
Publication status	Published - 2021 Sep 27
Event	26th International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2021 - Dallas, United States Duration: 2021 Sep 27 → 2021 Sep 29

Publication series

Name	International Conference on Simulation of Semiconductor Processes and Devices, SISPAD
Volume	2021-September

Conference

Conference	26th International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2021
Country/Territory	United States
City	Dallas
Period	21/9/27 → 21/9/29

Bibliographical note

Funding Information:
ACKNOWLEDGEMENT This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (NRF-2019R1A2C1086656 and NRF-2020M3H4A3081800). This work was also supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIT) (No.2019-0-01351, Development of Ultra Low-Power Mobile Deep Learning Semiconductor with Compression/Decompression of Activation/Kernel Data, 30 %)

Publisher Copyright:
© 2021 IEEE.

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering
Computer Science Applications
Modelling and Simulation

Access to Document

10.1109/SISPAD54002.2021.9592540

Cite this

Han, S. C., Choi, J., & Hong, S. M. (2021). Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network. In SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings (pp. 52-55). (International Conference on Simulation of Semiconductor Processes and Devices, SISPAD; Vol. 2021-September). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SISPAD54002.2021.9592540

Han, Seung Cheol ; Choi, Jonghyun ; Hong, Sung Min. / Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network. SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 52-55 (International Conference on Simulation of Semiconductor Processes and Devices, SISPAD).

@inproceedings{7e14b244c0b34e8da88a7a3ca0a7a7e7,

title = "Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network",

abstract = "We propose to use a 3D convolutional neural network to accelerate three-dimensional device simulation by generating an electrostatic potential profile. In the training phase, the deep neural network is trained with the simulation results for various 3D MOSFETs in a supervised manner. The generated potential profile is used as an initial guess at a non-equilibrium condition, while carrier densities are estimated by the frozen field simulation. By numerical examples for three-dimensional MOSFETs, we show that the proposed method significantly reduces the number of the Newton iterations. ",

author = "Han, {Seung Cheol} and Jonghyun Choi and Hong, {Sung Min}",

note = "Funding Information: ACKNOWLEDGEMENT This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (NRF-2019R1A2C1086656 and NRF-2020M3H4A3081800). This work was also supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIT) (No.2019-0-01351, Development of Ultra Low-Power Mobile Deep Learning Semiconductor with Compression/Decompression of Activation/Kernel Data, 30 %) Publisher Copyright: {\textcopyright} 2021 IEEE.; 26th International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2021 ; Conference date: 27-09-2021 Through 29-09-2021",

year = "2021",

month = sep,

day = "27",

doi = "10.1109/SISPAD54002.2021.9592540",

language = "English",

series = "International Conference on Simulation of Semiconductor Processes and Devices, SISPAD",

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

pages = "52--55",

booktitle = "SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings",

address = "United States",

}

Han, SC, Choi, J & Hong, SM 2021, Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network. in SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings. International Conference on Simulation of Semiconductor Processes and Devices, SISPAD, vol. 2021-September, Institute of Electrical and Electronics Engineers Inc., pp. 52-55, 26th International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2021, Dallas, United States, 21/9/27. https://doi.org/10.1109/SISPAD54002.2021.9592540

Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network. / Han, Seung Cheol; Choi, Jonghyun; Hong, Sung Min.

SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. p. 52-55 (International Conference on Simulation of Semiconductor Processes and Devices, SISPAD; Vol. 2021-September).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network

AU - Han, Seung Cheol

AU - Choi, Jonghyun

AU - Hong, Sung Min

N1 - Funding Information: ACKNOWLEDGEMENT This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (NRF-2019R1A2C1086656 and NRF-2020M3H4A3081800). This work was also supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIT) (No.2019-0-01351, Development of Ultra Low-Power Mobile Deep Learning Semiconductor with Compression/Decompression of Activation/Kernel Data, 30 %) Publisher Copyright: © 2021 IEEE.

PY - 2021/9/27

Y1 - 2021/9/27

N2 - We propose to use a 3D convolutional neural network to accelerate three-dimensional device simulation by generating an electrostatic potential profile. In the training phase, the deep neural network is trained with the simulation results for various 3D MOSFETs in a supervised manner. The generated potential profile is used as an initial guess at a non-equilibrium condition, while carrier densities are estimated by the frozen field simulation. By numerical examples for three-dimensional MOSFETs, we show that the proposed method significantly reduces the number of the Newton iterations.

AB - We propose to use a 3D convolutional neural network to accelerate three-dimensional device simulation by generating an electrostatic potential profile. In the training phase, the deep neural network is trained with the simulation results for various 3D MOSFETs in a supervised manner. The generated potential profile is used as an initial guess at a non-equilibrium condition, while carrier densities are estimated by the frozen field simulation. By numerical examples for three-dimensional MOSFETs, we show that the proposed method significantly reduces the number of the Newton iterations.

UR - http://www.scopus.com/inward/record.url?scp=85119409461&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85119409461&partnerID=8YFLogxK

U2 - 10.1109/SISPAD54002.2021.9592540

DO - 10.1109/SISPAD54002.2021.9592540

M3 - Conference contribution

AN - SCOPUS:85119409461

T3 - International Conference on Simulation of Semiconductor Processes and Devices, SISPAD

SP - 52

EP - 55

BT - SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 26th International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2021

Y2 - 27 September 2021 through 29 September 2021

ER -

Han SC, Choi J, Hong SM. Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network. In SISPAD 2021 - 2021 International Conference on Simulation of Semiconductor Processes and Devices, Proceedings. Institute of Electrical and Electronics Engineers Inc. 2021. p. 52-55. (International Conference on Simulation of Semiconductor Processes and Devices, SISPAD). https://doi.org/10.1109/SISPAD54002.2021.9592540

Acceleration of Three-Dimensional Device Simulation with the 3D Convolutional Neural Network

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this