Variability Aware FET Model With Physics Knowledge Based Machine Learning

We present variability-aware, computationally efficient, models for Fin Field Effect Transistors (FinFETs) using various machine learning (ML) architectures. This paper provides a detailed comparison of the various architectures. Our physics knowledge-based artificial neural networks (ANNs) demonstr...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	2023 7th IEEE Electron Devices Technology & Manufacturing Conference (EDTM) S. 1 - 3
Hauptverfasser:	Sheelvardhan, Kumar, Guglani, Surila, Ehteshamuddin, M., Roy, Sourajeet, Dasgupta, Avirup
Format:	Tagungsbericht
Sprache:	Englisch
Veröffentlicht:	IEEE 07.03.2023
Schlagworte:	Artificial neural networks Computational modeling Computer architecture Field effect transistors Knowledge based systems Machine learning Training
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	We present variability-aware, computationally efficient, models for Fin Field Effect Transistors (FinFETs) using various machine learning (ML) architectures. This paper provides a detailed comparison of the various architectures. Our physics knowledge-based artificial neural networks (ANNs) demonstrate unprecedented modeling efficiency. This is the first work presenting Prior Knowledge with Input Difference (PKID) ANN architecture for device modeling.
DOI:	10.1109/EDTM55494.2023.10103099