2022
DOI: 10.1109/tcad.2022.3166103
|View full text |Cite
|
Sign up to set email alerts
|

A Space–Time Neural Network for Analysis of Stress Evolution Under DC Current Stressing

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
2
0

Year Published

2022
2022
2023
2023

Publication Types

Select...
4
2

Relationship

1
5

Authors

Journals

citations
Cited by 8 publications
(2 citation statements)
references
References 40 publications
0
2
0
Order By: Relevance
“…One limitation of employing these competing learningbased schemes in EM analysis is that the prediction accuracy will decrease as the number of segments increases since the methods focus on solving a single PDE and cannot directly provide a global approximation for the interconnects governed by coupled PDEs subject to complex BCs. To mitigate this problem, our previous work [25] extended PINN to a new space-time physics-informed neural network (STPINN) for analyzing the EM-induced stress evolution by coupling the physics-based EM analysis with dynamic temperature incorporating Joule heating and via effect.…”
Section: Interconnect Treementioning
confidence: 99%
“…One limitation of employing these competing learningbased schemes in EM analysis is that the prediction accuracy will decrease as the number of segments increases since the methods focus on solving a single PDE and cannot directly provide a global approximation for the interconnects governed by coupled PDEs subject to complex BCs. To mitigate this problem, our previous work [25] extended PINN to a new space-time physics-informed neural network (STPINN) for analyzing the EM-induced stress evolution by coupling the physics-based EM analysis with dynamic temperature incorporating Joule heating and via effect.…”
Section: Interconnect Treementioning
confidence: 99%
“…Chen et al (2021) investigated the fatigue life of wafer-level chip scale packaging (WLCSP) using a hybrid finite element-neural network approach and developed a feasible finite element-artificial neural network (FEM-ANN) hybrid method to solve finite element problems in batch. Hou et al (2022) proposed a novel meshless model for computing stress evolution induced by electromigration in ultra-large-scale integrated circuits, which eliminates the need for time discretization and grid generation in traditional numerical stress evolution analysis, thereby saving computation time while ensuring satisfactory accuracy. Zippelius et al (2022) investigated the thermal-mechanical fatigue of different Sn-Ag-Cu (SAC) solders by transient thermal analysis (TTA), and predicted their thermal-mechanical fatigue behavior using ANNs.…”
Section: Introductionmentioning
confidence: 99%