Yuxuan Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yuxuan Zhu has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Kai Xu
The College of Integrated Circuits, Zhejiang University, Hangzhou, China
2 shared publications Yitao Ma
College of integrated circuits, Zhejiang University, Hangzhou, Zhejiang, China
2 shared publicationsChenggang Xu
College of integrated circuits, Zhejiang University, Hangzhou, Zhejiang, China
1 shared publication- Dawei Gao
Zhejiang University
1 shared publication Hongyu Tang
College of integrated circuits, Zhejiang University, Hangzhou, Zhejiang, China
1 shared publication- Hongyu Tang
College of integrated circuits, Zhejiang University, Hangzhou, Zhejiang, China
1 shared publication
Latest Publications
- 18 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72 (9) , 5089-5096
Meta-Optimized LDMOS Process–Device Co-Optimization: Extrapolation-Enhanced Modeling With Wafer-Level Validation
Accurate and efficient modeling of lateral double-diffused MOS (LDMOS) devices is critical for process optimization and reliability analysis, especially under limited simulation budgets. However, data-driven modeling for semiconductor devices faces three key challenges: limited training data due to the high cost of technology computer-aided design (TCAD) and silicon measurements; poor generalization to extrapolated or extreme configurations; and a gap between simulation and measured data, which undermines predictive reliability in practical use cases. To address these issues, this work presents a model-agnostic meta-learning (MAML) framework that improves the prediction accuracy and adaptability of deep neural networks and convolutional neural networks (CNNs) for TCAD-based process–device modeling. To further evaluate generalization to extrapolated scenarios, we introduce an additional dataset of samples whose input parameters lie within the design space but whose electrical outputs are near or beyond the original training range. This setup mimics real-world edge cases where accurate prediction is essential for safe operating area (SOA) design. Subsequently, k-means clustering is used to define distinct tasks, enabling task-specific fine-tuning. MAML-based models show significant performance improvements under this configuration, approaching in-distribution accuracy. Moreover, wafer-level experiments further validate the model’s guidance, confirming improvements in breakdown voltage without compromising on-resistance. Taken together, these results demonstrate that the proposed MAML framework effectively enhances model generalization and sample efficiency in semiconductor device modeling, supporting scalable and robust prediction under data-limited and distribution-shifted conditions.
- 18 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72 (9) , 5089-5096
Meta-Optimized LDMOS Process–Device Co-Optimization: Extrapolation-Enhanced Modeling With Wafer-Level Validation
Accurate and efficient modeling of lateral double-diffused MOS (LDMOS) devices is critical for process optimization and reliability analysis, especially under limited simulation budgets. However, data-driven modeling for semiconductor devices faces three key challenges: limited training data due to the high cost of technology computer-aided design (TCAD) and silicon measurements; poor generalization to extrapolated or extreme configurations; and a gap between simulation and measured data, which undermines predictive reliability in practical use cases. To address these issues, this work presents a model-agnostic meta-learning (MAML) framework that improves the prediction accuracy and adaptability of deep neural networks and convolutional neural networks (CNNs) for TCAD-based process–device modeling. To further evaluate generalization to extrapolated scenarios, we introduce an additional dataset of samples whose input parameters lie within the design space but whose electrical outputs are near or beyond the original training range. This setup mimics real-world edge cases where accurate prediction is essential for safe operating area (SOA) design. Subsequently, k-means clustering is used to define distinct tasks, enabling task-specific fine-tuning. MAML-based models show significant performance improvements under this configuration, approaching in-distribution accuracy. Moreover, wafer-level experiments further validate the model’s guidance, confirming improvements in breakdown voltage without compromising on-resistance. Taken together, these results demonstrate that the proposed MAML framework effectively enhances model generalization and sample efficiency in semiconductor device modeling, supporting scalable and robust prediction under data-limited and distribution-shifted conditions.
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Electron Device Letters
- Vol. 46 (9) , 1593-1596
Efficient LDMOS Design via Transferable Surrogate Models and Multi-Objective Optimization
Optimizing LDMOS performance requires balancing breakdown voltage (BV) and specific on-resistance ( $\text {R}_{\text {on},\text {sp}}$ ) under silicon-limit constraints. Conventional technology computer-aided design (TCAD)-based device design is time-consuming and inefficient for large parameter spaces. This work presents a machine learning (ML)-assisted framework that combines initial and fine-tuned deep neural network (DNN) surrogate models with multi-objective particle swarm optimization (MOPSO). The fine-tuned DNN adapts to a non-overlapping extended design space using only a small dataset, while the two surrogates are selectively applied during MOPSO to evaluate candidate designs, enabling significantly faster design evaluation compared to TCAD. SHAP analysis reveals consistent feature importance that aligns with the underlying device physics. The framework constructs diverse Pareto-optimal fronts, offering a scalable solution for automated LDMOS optimization under complex performance trade-offs.
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Electron Device Letters
- Vol. 46 (9) , 1593-1596
Efficient LDMOS Design via Transferable Surrogate Models and Multi-Objective Optimization
Optimizing LDMOS performance requires balancing breakdown voltage (BV) and specific on-resistance ( $\text {R}_{\text {on},\text {sp}}$ ) under silicon-limit constraints. Conventional technology computer-aided design (TCAD)-based device design is time-consuming and inefficient for large parameter spaces. This work presents a machine learning (ML)-assisted framework that combines initial and fine-tuned deep neural network (DNN) surrogate models with multi-objective particle swarm optimization (MOPSO). The fine-tuned DNN adapts to a non-overlapping extended design space using only a small dataset, while the two surrogates are selectively applied during MOPSO to evaluate candidate designs, enabling significantly faster design evaluation compared to TCAD. SHAP analysis reveals consistent feature importance that aligns with the underlying device physics. The framework constructs diverse Pareto-optimal fronts, offering a scalable solution for automated LDMOS optimization under complex performance trade-offs.