Application of Deep Artificial Neural Network to Model Characteristic Fluctuation of Multi-Channel Gate-All-Around Silicon Nanosheet and Nanofin MOSFETs Induced by Random Nanosized Metal Grains

Sagarika Dash, Yiming Li*, Wen Li Sung

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

In this work, we propose a deep artificial neural network (D-ANN) to estimate the work function fluctuation (WKF) on 4-channel stacked gate-all-around (GAA) silicon (Si) nanosheet (NS) and nanofin (NF) MOSFET devices for the first time. The 2-layered simple deep model can well predict the transfer characteristics for both NS/NF FET with a large number of (128) input features, utilizing considerably lesser (1100 samples) data uniformly. The resultant model is evaluated by the R2 score and RMSE to witness its competency and the average error is < 4%. We do also discuss the circuit simulation possibility by applying the ANN approach.

Original languageEnglish
Title of host publication7th IEEE Electron Devices Technology and Manufacturing Conference
Subtitle of host publicationStrengthen the Global Semiconductor Research Collaboration After the Covid-19 Pandemic, EDTM 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798350332520
DOIs
StatePublished - 2023
Event7th IEEE Electron Devices Technology and Manufacturing Conference, EDTM 2023 - Seoul, Korea, Republic of
Duration: 7 Mar 202310 Mar 2023

Publication series

Name7th IEEE Electron Devices Technology and Manufacturing Conference: Strengthen the Global Semiconductor Research Collaboration After the Covid-19 Pandemic, EDTM 2023

Conference

Conference7th IEEE Electron Devices Technology and Manufacturing Conference, EDTM 2023
Country/TerritoryKorea, Republic of
CitySeoul
Period7/03/2310/03/23

Keywords

  • artificial neural network
  • deep learning
  • gate all around
  • metal side wall
  • nanofin
  • nanosheet

Fingerprint

Dive into the research topics of 'Application of Deep Artificial Neural Network to Model Characteristic Fluctuation of Multi-Channel Gate-All-Around Silicon Nanosheet and Nanofin MOSFETs Induced by Random Nanosized Metal Grains'. Together they form a unique fingerprint.

Cite this