Development of a method for dynamic IR drop analysis based on machine learning
Informacionnye Tehnologii · 20.11.2026 · 32 (11) · 574-582
DOI: 10.17587/it.32.574-582 ↗Kurz gefasst
Fachartikel in der Zeitschrift Informacionnye Tehnologii, online erschienen am 20.11.2026. Verfasst von E. D. Demidov und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 22 Literaturangaben.
Abstract (Originalsprache)
This paper presents the development of a machine learning-based method for dynamic IR drop analysis. The proposed approach includes a methodology for training a neural network model using circuit characteristics, as well as a method for generating the input data supplied to the machine learning model. The paper describes the selected neural network architecture, the evaluation metrics used during training, and a customized loss function designed to improve the accuracy of dynamic IR drop prediction. Validation results obtained on an independent test dataset are presented, and the proposed approach is evaluated in terms of prediction accuracy and execution time compared with the conventional dynamic IR drop analysis methodology. The validation results demonstrate that the proposed model achieves a mean absolute error (MAE) of less than 400 mV and a mean relative error below 20 %. Furthermore, the developed approach provides a significant reduction in dynamic IR drop analysis time compared with the conventional method. As the analyzed circuit grows in size, the achieved speedup steadily increases.
Zitieren
APA
Demidov, E. D., & Vasilev, A. A. (2026). Development of a method for dynamic IR drop analysis based on machine learning. Informacionnye Tehnologii, 32(11), 574-582. https://doi.org/10.17587/it.32.574-582
Vancouver
Demidov ED, Vasilev AA. Development of a method for dynamic IR drop analysis based on machine learning. Informacionnye Tehnologii. 2026;32(11):574-582. doi:10.17587/it.32.574-582
BibTeX
@article{Demidov2026,
author = {Demidov, E. D. and Vasilev, A. A.},
title = {Development of a method for dynamic IR drop analysis based on machine learning},
journal = {Informacionnye Tehnologii},
year = {2026},
volume = {32},
number = {11},
pages = {574--582},
doi = {10.17587/it.32.574-582}
}
EUVIMED