Physics-aware Jacobian calibration for bridging the simulation-to-real gap in LDMOS modeling
The simulation-to-real gap between idealized Technology computer-aided design (TCAD) simulations and experimental data from fabricated devices limits the accuracy of data-driven modeling for semiconductor devices. This work proposes a Physics-Aware Jacobian Calibration framework to improve neural network surrogate modeling of laterally diffused metal oxide semiconductor (LDMOS) devices under sparse experimental supervision. A physics prior model trained on large-scale TCAD simulations provides first order response sensitivities through its Jacobian matrix, enabling corrections from a small set of experimental anchor samples to propagate across the design space. Using the corrected simulation data together with experimental samples, a calibrated proxy model (CPM) is constructed for device performance prediction. Experimental results on LDMOS devices show that the CPM achieves an average mean absolute percentage error (MAPE) of 2.28% on an unseen experimental test set while preserving physically consistent trends between design parameters and device performance. The proposed framework provides a scalable approach for physics informed AI assisted semiconductor modeling under realistic process variations.
Authors
- K Z Wang (ORCID: https://orcid.org/0009-0005-1329-1219)
- Feng Yan (ORCID: https://orcid.org/0009-0008-0178-0594)
- Tianfei Lei
- Zerui Jin (ORCID: https://orcid.org/0009-0005-0686-6037)
- Zhihao Li (ORCID: https://orcid.org/0000-0002-2066-8775)
- Wenbin Zhou (ORCID: https://orcid.org/0009-0006-5998-4456)
- Wei Si
- Tiantian Shi (ORCID: https://orcid.org/0009-0007-7314-010X)
- Yiming Liao
- Xiao Li Ji
Institutions
- Nanjing University of Science and Technology (CN)
- Nanjing University (CN)
Publication Details
- Journal
- Journal of Physics D Applied Physics
- Published
- 2026-11-19
- DOI
- https://doi.org/10.1088/1361-6463/aebe3b
- Primary Topic
- Advancements in Semiconductor Devices and Circuit Design
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China
- Nanjing University