Deep learning-based ASM-HEMT I-V parameter extraction

Fredo Chavez; Devin T. Davis; Nicholas C. Miller; Sourabh Khandelwal

doi:10.1109/LED.2022.3197800

Deep learning-based ASM-HEMT I-V parameter extraction

Fredo Chavez^*, Devin T. Davis, Nicholas C. Miller, Sourabh Khandelwal

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

15 Citations (Scopus)

Abstract

A fast and accurate deep learning (DL) based ASM-HEMT I-V model parameter extraction is presented for the first time. DL-based extraction starts with 120k training data-sets comprising of 374 million I-V data points. Training data-sets are generated through Monte Carlo simulations. The trained DL-model is demonstrated to successfully model 114 GaN HEMTs from a typical GaN fabrication process. The predicted parameters show an excellent fit for the I-V data. In addition, the root-mean-square(RMS) error incurred for key electrical parameters such as pinch-off voltage, linear condition current and the maximum current is 2.2%, 17.6%, and 2.4% respectively. The proposed approach is verified for multiple GaN HEMTs of different sizes. The developed technique can provide a very fast means for parameter extraction with a reasonable accuracy.

Original language	English
Pages (from-to)	1633-1636
Number of pages	4
Journal	IEEE Electron Device Letters
Volume	43
Issue number	10
DOIs	https://doi.org/10.1109/LED.2022.3197800
Publication status	Published - Oct 2022

Access to Document

10.1109/LED.2022.3197800

Cite this

@article{6991ac37112e4035ac19c6702f39ce47,

title = "Deep learning-based ASM-HEMT I-V parameter extraction",

abstract = "A fast and accurate deep learning (DL) based ASM-HEMT I-V model parameter extraction is presented for the first time. DL-based extraction starts with 120k training data-sets comprising of 374 million I-V data points. Training data-sets are generated through Monte Carlo simulations. The trained DL-model is demonstrated to successfully model 114 GaN HEMTs from a typical GaN fabrication process. The predicted parameters show an excellent fit for the I-V data. In addition, the root-mean-square(RMS) error incurred for key electrical parameters such as pinch-off voltage, linear condition current and the maximum current is 2.2%, 17.6%, and 2.4% respectively. The proposed approach is verified for multiple GaN HEMTs of different sizes. The developed technique can provide a very fast means for parameter extraction with a reasonable accuracy.",

author = "Fredo Chavez and Davis, {Devin T.} and Miller, {Nicholas C.} and Sourabh Khandelwal",

year = "2022",

month = oct,

doi = "10.1109/LED.2022.3197800",

language = "English",

volume = "43",

pages = "1633--1636",

journal = "IEEE Electron Device Letters",

issn = "0741-3106",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

number = "10",

}

TY - JOUR

T1 - Deep learning-based ASM-HEMT I-V parameter extraction

AU - Chavez, Fredo

AU - Davis, Devin T.

AU - Miller, Nicholas C.

AU - Khandelwal, Sourabh

PY - 2022/10

Y1 - 2022/10

N2 - A fast and accurate deep learning (DL) based ASM-HEMT I-V model parameter extraction is presented for the first time. DL-based extraction starts with 120k training data-sets comprising of 374 million I-V data points. Training data-sets are generated through Monte Carlo simulations. The trained DL-model is demonstrated to successfully model 114 GaN HEMTs from a typical GaN fabrication process. The predicted parameters show an excellent fit for the I-V data. In addition, the root-mean-square(RMS) error incurred for key electrical parameters such as pinch-off voltage, linear condition current and the maximum current is 2.2%, 17.6%, and 2.4% respectively. The proposed approach is verified for multiple GaN HEMTs of different sizes. The developed technique can provide a very fast means for parameter extraction with a reasonable accuracy.

AB - A fast and accurate deep learning (DL) based ASM-HEMT I-V model parameter extraction is presented for the first time. DL-based extraction starts with 120k training data-sets comprising of 374 million I-V data points. Training data-sets are generated through Monte Carlo simulations. The trained DL-model is demonstrated to successfully model 114 GaN HEMTs from a typical GaN fabrication process. The predicted parameters show an excellent fit for the I-V data. In addition, the root-mean-square(RMS) error incurred for key electrical parameters such as pinch-off voltage, linear condition current and the maximum current is 2.2%, 17.6%, and 2.4% respectively. The proposed approach is verified for multiple GaN HEMTs of different sizes. The developed technique can provide a very fast means for parameter extraction with a reasonable accuracy.

UR - http://www.scopus.com/inward/record.url?scp=85136068227&partnerID=8YFLogxK

U2 - 10.1109/LED.2022.3197800

DO - 10.1109/LED.2022.3197800

M3 - Article

AN - SCOPUS:85136068227

SN - 0741-3106

VL - 43

SP - 1633

EP - 1636

JO - IEEE Electron Device Letters

JF - IEEE Electron Device Letters

IS - 10

ER -

Deep learning-based ASM-HEMT I-V parameter extraction

Abstract

Access to Document

Other files and links

Fingerprint

Cite this