Embedding-Enhanced Probabilistic Modeling of Ferroelectric Field Effect Transistors (FeFETs)

TL;DR

This paper introduces a probabilistic modeling framework for FeFETs that captures device variability using a Mixture Density Network with embedding layers, enabling more accurate and stable circuit simulations.

Contribution

It presents a novel embedding-enhanced probabilistic model based on MDNs with smooth activation functions for variability-aware FeFET modeling.

Findings

Achieved R2 of 0.92 in modeling FeFET variability

Enabled generation of synthetic device instances for simulation

Provided a scalable, data-driven modeling approach

Abstract

FeFETs hold strong potential for advancing memory and logic technologies, but their inherent randomness arising from both operational cycling and fabrication variability poses significant challenges for accurate and reliable modeling. Capturing this variability is critical, as it enables designers to predict behavior, optimize performance, and ensure reliability and robustness against variations in manufacturing and operating conditions. Existing deterministic and machine learning-based compact models often fail to capture the full extent of this variability or lack the mathematical smoothness required for stable circuit-level integration. In this work, we present an enhanced probabilistic modeling framework for FeFETs that addresses these limitations. Building upon a Mixture Density Network (MDN) foundation, our approach integrates C-infinity continuous activation functions for smooth, stable learning and a device-specific embedding layer to capture intrinsic physical variability across devices. Sampling from the learned embedding distribution enables the generation of synthetic device instances for variability-aware simulation. With an R2 of 0.92, the model demonstrates high accuracy in capturing the variability of FeFET current behavior. Altogether, this framework provides a scalable, data-driven solution for modeling the full stochastic behavior of FeFETs and offers a strong foundation for future compact model development and circuit simulation integration.