An Overview of AI-Assisted Design-on-Simulation Technology for Reliability Life Prediction of Advanced Packaging

Abstract

Several design parameters affect the reliability of wafer-level type advanced packaging, such as upper and lower pad sizes, solder volume, buffer layer thickness, and chip thickness, etc. Conventionally, the accelerated thermal cycling test (ATCT) is used to evaluate the reliability life of electronic packaging; however, optimizing the design parameters through ATCT is time-consuming and expensive, reducing the number of experiments becomes a critical issue. In recent years, many researchers have adopted the finite-element-based design-on-simulation (DoS) technology for the reliability assessment of electronic packaging. DoS technology can effectively shorten the design cycle, reduce costs, and effectively optimize the packaging structure. However, the simulation analysis results are highly dependent on the individual researcher and are usually inconsistent between them. Artificial intelligence (AI) can help researchers avoid the shortcomings of the human factor. This study demonstrates AI-assisted DoS technology by combining artificial intelligence and simulation technologies to predict wafer level package (WLP) reliability. In order to ensure reliability prediction accuracy, the simulation procedure was validated by several experiments prior to creating a large AI training database. This research studies several machine learning models, including artificial neural network (ANN), recurrent neural network (RNN), support vector regression (SVR), kernel ridge regression (KRR), K-nearest neighbor (KNN), and random forest (RF). These models are evaluated in this study based on prediction accuracy and CPU time consumption.

Predicting Wafer-Level Package Reliability Life Using Mixed Supervised and Unsupervised Machine Learning Algorithms

“…In the case of WLP, the thermal loading failure usually occurs at the outermost diagonal solder ball of the package. This study uses five WLP test vehicles (TV: WLP1-5) [ 16 , 17 , 18 ] and one fan-out WLP (FO-WLP, [ 19 ]) for FEA model validation. The structure component sizes, materials, and mean-time-to-failure (MTTF) reliability life are shown in Table 1 , Table 2 and Table 3 [ 17 , 18 , 19 ].…”

Development of ANN-Based Warpage Prediction Model for FCCSP via Subdomain Sampling and Taguchi Hyperparameter Optimization

“…The NN prediction model performance can be alternatively improved through hyperparameter tuning [ 21 , 22 , 23 , 24 , 25 ]. Hyperparameters are crucial for the performance of a machine learning model because they control the architecture of a neural network.…”

“…Well-tuned hyperparameters can also prevent the model from overfitting or underfitting (see, e.g., [ 26 ]). In the literature, the hyperparameters were mostly tuned using trial-and-error parametric analysis (one factor at a time) [ 21 ], grid search [ 22 ], and random search [ 23 ]. The former two approaches are either unable to account for the interaction effect of hyperparameters, or computationally expensive, especially for models with a large number of hyperparameters and a huge search space.…”

Study on the Strip Warpage Issues Encountered in the Flip-Chip Process

“…Yao et al [ 18 ] proposed an analytical model to evaluate the pore and superficial permeability of an underfill porous medium in a flip-chip packaging; they also presented an approach to predict the flow front and the filling time [ 19 ]. Chiang et al [ 20 ] proposed an overview of artificial intelligence assisted design on simulation technology for reliability life prediction of advanced packaging. Developers only need to input geometric data of the package structures, and then the reliability life cycle can be obtained by this AI-trained model.…”

“…Severe strip warpage occurring during the post-molding period reduces product yield. The literature [ 8 , 9 , 10 , 11 , 12 , 13 , 14 , 20 , 21 , 22 ] indicates that FE simulation is frequently used to solve packaging warpage problems. Material parameters are crucial for determining whether consistency between simulated and experimental values is achieved.…”