Yiliang Guo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yiliang Guo has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Madhavan Swaminathan
The Pennsylvania State University
3 shared publications - Yifan Wang
School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA
3 shared publications - Joshua Corsello
Epirus Inc, Torrance, CA, USA
2 shared publications - Xingchen Li
School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA
2 shared publications Madhavan Swaminathan
School of Electrical Engineering and Computer Science, Pennsylvania State University, State College, PA, USA
1 shared publicationRahul Kumar
School of Electrical Engineering and Computer Science, Pennsylvania State University, State College, PA, USA
1 shared publication
Latest Publications
- 2 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 11, 291-300
Vision-Based Transformer Model for Fast Simulation of Advanced Packaging Structure
Accurate electromagnetic (EM) modeling of electronic package structures is critical for signal and power integrity analysis, yet full-wave simulations on fine meshes are often computationally expensive. Deep learning and neural network models have recently attracted considerable attention as alternative solutions. However, with the continuous expansion of the design space in advanced packaging, conventional surrogate architectures such as FCNNs, CNNs, and RNNs start to exhibit limitations in scalability. Therefore, in this paper, we propose two vision-based transformer models that eliminate the need for design parameters and leverage self-attention mechanism for efficient package-level EM and multi-physics simulation. The first framework employs a Vision Transformer (ViT) to directly predict frequency-domain responses (S-parameters) from structural images of the package. By capturing global spatial dependencies through self-attention, the ViT-based model achieves high accuracy while significantly reducing computational cost compared to conventional EM solvers. The second framework introduces a transformer-based super-resolution neural network designed for multiphysics co-simulation scenarios. Specifically, a physical structure is simulated with a coarse mesh using an EM solver to provide power input to the thermal solver, which subsequently generates low-resolution temperature profile images. The super-resolution neural network then refines these results to produce high-resolution temperature distributions, akin to those obtained through numerical simulations with a fine mesh. Various numerical examples are simulated to validate the capability and improvement of the proposed method.
- 2 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 11, 291-300
Vision-Based Transformer Model for Fast Simulation of Advanced Packaging Structure
Accurate electromagnetic (EM) modeling of electronic package structures is critical for signal and power integrity analysis, yet full-wave simulations on fine meshes are often computationally expensive. Deep learning and neural network models have recently attracted considerable attention as alternative solutions. However, with the continuous expansion of the design space in advanced packaging, conventional surrogate architectures such as FCNNs, CNNs, and RNNs start to exhibit limitations in scalability. Therefore, in this paper, we propose two vision-based transformer models that eliminate the need for design parameters and leverage self-attention mechanism for efficient package-level EM and multi-physics simulation. The first framework employs a Vision Transformer (ViT) to directly predict frequency-domain responses (S-parameters) from structural images of the package. By capturing global spatial dependencies through self-attention, the ViT-based model achieves high accuracy while significantly reducing computational cost compared to conventional EM solvers. The second framework introduces a transformer-based super-resolution neural network designed for multiphysics co-simulation scenarios. Specifically, a physical structure is simulated with a coarse mesh using an EM solver to provide power input to the thermal solver, which subsequently generates low-resolution temperature profile images. The super-resolution neural network then refines these results to produce high-resolution temperature distributions, akin to those obtained through numerical simulations with a fine mesh. Various numerical examples are simulated to validate the capability and improvement of the proposed method.
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. PP (21563950) , 1
Deep Kernel Based Hyperparameter Adaptive Learning and Frequency Response Predictions Using Transposed Convolutional Neural Network
Electronic Design Automation (EDA) has unique challenges for addressing the design of systems for emerging applications due to the complexities involved, where multiple chiplets are integrated together on a heterogeneous platform. This challenge arises due to the long computational time required for simulation to capture all the necessary first order, second order and sometimes third order parasitic effects. Emerging machine learning (ML) and Gaussian process (GP) methods have helped expedite these processes. The deep kernel learning (DKL) model combines the structural properties of deep learning architectures with the non-parametric flexibility of kernel methods. It shows advantages by applying a GP with the corresponding kernel function to the final hidden layer of a deep neural network. However, DKL sometimes suffers from overfitting and scalability issues. In this paper, we propose an adaptive learning framework for S-parameter prediction, incorporating the spectral transposed convolutional neural network (S-TCNN) and DKL. The proposed model takes input parameters from the design space, up-samples them through transposed convolutional layers, and utilizes a GP kernel layer to approximate the desired kernel function. Additionally, the latent feature space adaptively compresses and extracts features from the input matrix, serving as a separate input parameter for the GP kernel layer. Further, we discuss the training strategy and model scalability. The proposed model is tested and evaluated using two advanced packaging examples. Results show a reduction in the number of hyperparameters by over 50% and approximately 40% improvements in loss and normalized mean-square-error (NMSE).
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. PP (21563950) , 1
Deep Kernel Based Hyperparameter Adaptive Learning and Frequency Response Predictions Using Transposed Convolutional Neural Network
Electronic Design Automation (EDA) has unique challenges for addressing the design of systems for emerging applications due to the complexities involved, where multiple chiplets are integrated together on a heterogeneous platform. This challenge arises due to the long computational time required for simulation to capture all the necessary first order, second order and sometimes third order parasitic effects. Emerging machine learning (ML) and Gaussian process (GP) methods have helped expedite these processes. The deep kernel learning (DKL) model combines the structural properties of deep learning architectures with the non-parametric flexibility of kernel methods. It shows advantages by applying a GP with the corresponding kernel function to the final hidden layer of a deep neural network. However, DKL sometimes suffers from overfitting and scalability issues. In this paper, we propose an adaptive learning framework for S-parameter prediction, incorporating the spectral transposed convolutional neural network (S-TCNN) and DKL. The proposed model takes input parameters from the design space, up-samples them through transposed convolutional layers, and utilizes a GP kernel layer to approximate the desired kernel function. Additionally, the latent feature space adaptively compresses and extracts features from the input matrix, serving as a separate input parameter for the GP kernel layer. Further, we discuss the training strategy and model scalability. The proposed model is tested and evaluated using two advanced packaging examples. Results show a reduction in the number of hyperparameters by over 50% and approximately 40% improvements in loss and normalized mean-square-error (NMSE).
- 22 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 10, 259-270
GPU Accelerated Matrix Solution Using Novel Preconditioner for Three Dimensional Laguerre-FDTD Method
Conventionally, the large sparse matrix equation ($Ax=b$) generated by the Laguerre-FDTD method is computed using direct matrix solvers, which is often numerically expensive and computationally slow. In this work, we demonstrate an innovative approach to replace direct matrix solver with an iterative algorithm for the Laguerre-FDTD method. A novel preconditioner, specifically targeted to improve the convergence rate of biconjugate gradient stabilized solver (BiCGSTAB), is derived and implemented in the Laguerre-FDTD method. Compared with the classical Jacobi preconditioner, the proposed preconditioner achieves on average an improvement of more than 1.3× in the convergence rate. To further leverage the computational efficiency, a modified sparse matrix-vector multiplication algorithm is proposed and implemented using a General-Purpose Graphics Processing Unit (GPGPU). The new algorithm ensures that all computations are performed within the GPU, with minimum number of device-to-host data transfer and global memory access. With GPU's accelerated computing capability, the proposed solver achieves more than 5× computational speed up with respect to a high performance CPU-based direct solver on average. In addition, due to the intrinsic memory efficient nature of iterative solver, our approach also shows maximally more than 31× reduction in memory consumption against the direct solver. Various numerical examples are simulated to validate the capability and improvement of the proposed method.
- 22 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 10, 259-270
GPU Accelerated Matrix Solution Using Novel Preconditioner for Three Dimensional Laguerre-FDTD Method
Conventionally, the large sparse matrix equation ($Ax=b$) generated by the Laguerre-FDTD method is computed using direct matrix solvers, which is often numerically expensive and computationally slow. In this work, we demonstrate an innovative approach to replace direct matrix solver with an iterative algorithm for the Laguerre-FDTD method. A novel preconditioner, specifically targeted to improve the convergence rate of biconjugate gradient stabilized solver (BiCGSTAB), is derived and implemented in the Laguerre-FDTD method. Compared with the classical Jacobi preconditioner, the proposed preconditioner achieves on average an improvement of more than 1.3× in the convergence rate. To further leverage the computational efficiency, a modified sparse matrix-vector multiplication algorithm is proposed and implemented using a General-Purpose Graphics Processing Unit (GPGPU). The new algorithm ensures that all computations are performed within the GPU, with minimum number of device-to-host data transfer and global memory access. With GPU's accelerated computing capability, the proposed solver achieves more than 5× computational speed up with respect to a high performance CPU-based direct solver on average. In addition, due to the intrinsic memory efficient nature of iterative solver, our approach also shows maximally more than 31× reduction in memory consumption against the direct solver. Various numerical examples are simulated to validate the capability and improvement of the proposed method.
- 22 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. 13 (10) , 1533-1544
Extrapolation With Range Determination of 2-D Spectral Transposed Convolutional Neural Network for Advanced Packaging Problems
In this paper, we propose the application of 2D Spectral Transposed Convolutional Neural Network (S-TCNN) with extrapolation to reduce the number of trainable parameters during the up-sampling process, leading to reduced training time and decrease in computational resources. Our proposed model consists of three stages, namely the prediction stage that uses high accuracy 2D S-TCNN to predict partial frequency responses; the extrapolation stage that takes the output of prediction stage and uses Gaussian Process (GP) to construct the mean and covariance matrix; and the extrapolation range determining stage responses to numerically compute the optimized output length. For extrapolation, we construct new kernel combinations and use maximum likelihood estimation (MLE) to address the problem of losing correlation between priors and extrapolated points as distances increase. We also form a quantitative method to determine the extrapolation range, which takes the distance and confidence interval (CI) into consideration. We apply our method to three application examples: (1) Staggered via, (2) On-chip microstrip line and (3) 1×8 Slot antenna array. Results show that our model can reduce about 20% of the trainable parameters, 30% of memory storage and 20% of the training time compared with the 2D S-TCNN. The proposed method can also achieve a similar normalized mean-squared error (NMSE) level with a small trade-off in the final loss.
- 22 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Components, Packaging and Manufacturing Technology
- Vol. 13 (10) , 1533-1544
Extrapolation With Range Determination of 2-D Spectral Transposed Convolutional Neural Network for Advanced Packaging Problems
In this paper, we propose the application of 2D Spectral Transposed Convolutional Neural Network (S-TCNN) with extrapolation to reduce the number of trainable parameters during the up-sampling process, leading to reduced training time and decrease in computational resources. Our proposed model consists of three stages, namely the prediction stage that uses high accuracy 2D S-TCNN to predict partial frequency responses; the extrapolation stage that takes the output of prediction stage and uses Gaussian Process (GP) to construct the mean and covariance matrix; and the extrapolation range determining stage responses to numerically compute the optimized output length. For extrapolation, we construct new kernel combinations and use maximum likelihood estimation (MLE) to address the problem of losing correlation between priors and extrapolated points as distances increase. We also form a quantitative method to determine the extrapolation range, which takes the distance and confidence interval (CI) into consideration. We apply our method to three application examples: (1) Staggered via, (2) On-chip microstrip line and (3) 1×8 Slot antenna array. Results show that our model can reduce about 20% of the trainable parameters, 30% of memory storage and 20% of the training time compared with the 2D S-TCNN. The proposed method can also achieve a similar normalized mean-squared error (NMSE) level with a small trade-off in the final loss.