Machine learning based recursive partitioning for simplifying OPC model building complexity

Apoorva Oak; Soobin Hwang; Ruoxia Chen; Shinill Kang; Ryan Ryoung Han Kim

doi:10.1117/12.2584704

Machine learning based recursive partitioning for simplifying OPC model building complexity

Apoorva Oak, Soobin Hwang, Ruoxia Chen, Shinill Kang, Ryan Ryoung Han Kim

Department of Mechanical Engineering

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

Abstract

We propose the use of machine learning based analytics to simplify OPC (Optical Proximity Correction) model building process which demands concurrent optimization of more than 70 parameters as nodes shrink. We first built a deep neural network architecture to predict the RMS error, for a given set of model parameters. The neural network was trained on existing OPC model parameters and corresponding output RMS data of simulations to achieve an accurate prediction of output RMS for given set of OPC model parameters. Later, a sensitivity analysis-based methodology for recursive partitioning of OPC modelling parameters was employed to reduce the total search space of OPC model simulations. This resulted in reduction of the number of OPC model iterations performed during model tuning by orders of magnitude.

Original language	English
Title of host publication	Design-Process-Technology Co-optimization XV
Editors	Chi-Min Yuan, Ryoung-Han Kim
Publisher	SPIE
ISBN (Electronic)	9781510640610
DOIs	https://doi.org/10.1117/12.2584704
Publication status	Published - 2021
Event	Design-Process-Technology Co-optimization XV 2021 - Virtual, Online, United States Duration: 2021 Feb 22 → 2021 Feb 26

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	11614
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	Design-Process-Technology Co-optimization XV 2021
Country/Territory	United States
City	Virtual, Online
Period	21/2/22 → 21/2/26

Bibliographical note

Publisher Copyright:
© 2021 SPIE.

All Science Journal Classification (ASJC) codes

Electronic, Optical and Magnetic Materials
Condensed Matter Physics
Computer Science Applications
Applied Mathematics
Electrical and Electronic Engineering

Access to Document

10.1117/12.2584704

Cite this

Oak, A., Hwang, S., Chen, R., Kang, S., & Kim, R. R. H. (2021). Machine learning based recursive partitioning for simplifying OPC model building complexity. In C-M. Yuan, & R-H. Kim (Eds.), Design-Process-Technology Co-optimization XV [116140O] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11614). SPIE. https://doi.org/10.1117/12.2584704

@inproceedings{7815ceb33e194eb4b1906670c422a391,

title = "Machine learning based recursive partitioning for simplifying OPC model building complexity",

abstract = "We propose the use of machine learning based analytics to simplify OPC (Optical Proximity Correction) model building process which demands concurrent optimization of more than 70 parameters as nodes shrink. We first built a deep neural network architecture to predict the RMS error, for a given set of model parameters. The neural network was trained on existing OPC model parameters and corresponding output RMS data of simulations to achieve an accurate prediction of output RMS for given set of OPC model parameters. Later, a sensitivity analysis-based methodology for recursive partitioning of OPC modelling parameters was employed to reduce the total search space of OPC model simulations. This resulted in reduction of the number of OPC model iterations performed during model tuning by orders of magnitude.",

author = "Apoorva Oak and Soobin Hwang and Ruoxia Chen and Shinill Kang and Kim, {Ryan Ryoung Han}",

note = "Publisher Copyright: {\textcopyright} 2021 SPIE.; Design-Process-Technology Co-optimization XV 2021 ; Conference date: 22-02-2021 Through 26-02-2021",

year = "2021",

doi = "10.1117/12.2584704",

language = "English",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Chi-Min Yuan and Ryoung-Han Kim",

booktitle = "Design-Process-Technology Co-optimization XV",

address = "United States",

}

Oak, A, Hwang, S, Chen, R, Kang, S & Kim, RRH 2021, Machine learning based recursive partitioning for simplifying OPC model building complexity. in C-M Yuan & R-H Kim (eds), Design-Process-Technology Co-optimization XV., 116140O, Proceedings of SPIE - The International Society for Optical Engineering, vol. 11614, SPIE, Design-Process-Technology Co-optimization XV 2021, Virtual, Online, United States, 21/2/22. https://doi.org/10.1117/12.2584704

Machine learning based recursive partitioning for simplifying OPC model building complexity. / Oak, Apoorva; Hwang, Soobin; Chen, Ruoxia et al.

Design-Process-Technology Co-optimization XV. ed. / Chi-Min Yuan; Ryoung-Han Kim. SPIE, 2021. 116140O (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11614).

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

TY - GEN

T1 - Machine learning based recursive partitioning for simplifying OPC model building complexity

AU - Oak, Apoorva

AU - Hwang, Soobin

AU - Chen, Ruoxia

AU - Kang, Shinill

AU - Kim, Ryan Ryoung Han

PY - 2021

Y1 - 2021

N2 - We propose the use of machine learning based analytics to simplify OPC (Optical Proximity Correction) model building process which demands concurrent optimization of more than 70 parameters as nodes shrink. We first built a deep neural network architecture to predict the RMS error, for a given set of model parameters. The neural network was trained on existing OPC model parameters and corresponding output RMS data of simulations to achieve an accurate prediction of output RMS for given set of OPC model parameters. Later, a sensitivity analysis-based methodology for recursive partitioning of OPC modelling parameters was employed to reduce the total search space of OPC model simulations. This resulted in reduction of the number of OPC model iterations performed during model tuning by orders of magnitude.

AB - We propose the use of machine learning based analytics to simplify OPC (Optical Proximity Correction) model building process which demands concurrent optimization of more than 70 parameters as nodes shrink. We first built a deep neural network architecture to predict the RMS error, for a given set of model parameters. The neural network was trained on existing OPC model parameters and corresponding output RMS data of simulations to achieve an accurate prediction of output RMS for given set of OPC model parameters. Later, a sensitivity analysis-based methodology for recursive partitioning of OPC modelling parameters was employed to reduce the total search space of OPC model simulations. This resulted in reduction of the number of OPC model iterations performed during model tuning by orders of magnitude.

UR - http://www.scopus.com/inward/record.url?scp=85105435604&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85105435604&partnerID=8YFLogxK

U2 - 10.1117/12.2584704

DO - 10.1117/12.2584704

M3 - Conference contribution

AN - SCOPUS:85105435604

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Design-Process-Technology Co-optimization XV

A2 - Yuan, Chi-Min

A2 - Kim, Ryoung-Han

PB - SPIE

T2 - Design-Process-Technology Co-optimization XV 2021

Y2 - 22 February 2021 through 26 February 2021

ER -

Machine learning based recursive partitioning for simplifying OPC model building complexity

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this