Restructuring TCAD System: Teaching Traditional TCAD New Tricks

Myung, Sanghoon; Jang, Wonik; Jin, Shaohua; Choe, Jae Myung; Jeong, Changwook; Kim, Dae Sin

doi:10.1109/iedm19574.2021.9720616

Cited by 9 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From the deep learning model perspective, the data type of the former and latter are regarded as the real value (unstructured mesh data) and the two-dimensional (2D) image (meshless data), respectively. Hence, the previous works 12,18) have suggested the convolutional neural network (CNN)based deep learning models to predict the doping profile in accordance with those types. As a review paper, we will review and analyze the previous studies.…”

Section: Process Simulationmentioning

confidence: 99%

“…Figure 8 shows the doping profile on the unstructured mesh that the deep learning model typically cannot deal with. Thus, many previous works 13,[18][19][20][21] have suggested converting such profiles into a structured mesh, as shown in Fig. 8(b).…”

Section: Process Simulationmentioning

confidence: 99%

“…putting relative coordinate) described before because it cannot distinguish the size of grids. Hence, this work 18) has proposed a method, the so-called Real Time TCAD (RTT) process model, which samples the real coordinates from the structured mesh. Figure 9(a) illustrates the final architecture of the process model that applies the method.…”

Section: Process Simulationmentioning

confidence: 99%

See 2 more Smart Citations

Comprehensive studies on deep learning applicable to TCAD

Myung

Choi

Jang

et al. 2023

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

TCAD simulation has incessantly solved many complex problems, but it becomes demanding that alternatives be found because TCAD simulation cannot provide the precise and fast prediction in the nano-scale era. With the success story on deep learning in research area, many big data companies have attempted to introduce deep learning to support or replace TCAD simulation. The reason is deep learning models has great potential that solves the problems of the TCAD simulation in terms of execution time, coverage. This paper aims to describe various scenarios of deep learning applicable to TCAD. We firstly describe an application that supplies TCAD data to the deep learning model although TCAD simulation is not calibrated. We then review various approaches that mimic TCAD simulation itself. We finally introduce an application that deep learning model automatically calibrates TCAD models to the measurement without experts. In each scenario, we review the related papers and compare pros and cons.

show abstract

Section: Process Simulationmentioning

confidence: 99%

Section: Process Simulationmentioning

confidence: 99%

Section: Process Simulationmentioning

confidence: 99%

See 1 more Smart Citation

Comprehensive studies on deep learning applicable to TCAD

Myung

Choi

Jang

et al. 2023

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The dataset consists of the scalar values as an input and images as an output. We used the RTT model (Myung et al, 2020(Myung et al, , 2021b to solve this problem.…”

Section: Real-world Problemmentioning

confidence: 99%

PAC-Net: A Model Pruning Approach to Inductive Transfer Learning

Myung¹,

Huh²,

Jang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Inductive transfer learning aims to learn from a small amount of training data for the target task by utilizing a pre-trained model from the source task. Most strategies that involve large-scale deep learning models adopt initialization with the pretrained model and fine-tuning for the target task. However, when using over-parameterized models, we can often prune the model without sacrificing the accuracy of the source task. This motivates us to adopt model pruning for transfer learning with deep learning models. In this paper, we propose PAC-Net, a simple yet effective approach for transfer learning based on pruning. PAC-Net consists of three steps: Prune, Allocate, and Calibrate (PAC). The main idea behind these steps is to identify essential weights for the source task, fine-tune on the source task by updating the essential weights, and then calibrate on the target task by updating the remaining redundant weights. Under the various and extensive set of inductive transfer learning experiments, we show that our method achieves state-of-the-art performance by a large margin.

show abstract

“…In various simulation fields governed by partial differential equations (PDEs), such as TCAD, computational fluid dynamics, and structural analysis, methodologies to model the relationship between the simulation parameters and its output using machine learning rather than using conventional solvers based on the difference method are being actively investigated. [6][7][8][9][10][11][12][13][14][15] For example, for the simulation of semiconductors, neural networks (NNs) have been used to predict device characteristics of metaloxide-semiconductor field-effect transistors (MOSFETs) [16,17] and Gaussian process regression has been used for GaN lightemitting diode structure optimization. [18] The annealing process for recovering crystals containing point defects introduced by ion implantation has a significant impact on the device properties.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning for Semiconductor Process Simulation Described by Coupled Partial Differential Equations

Sato

Kutsukake

Harada

et al. 2023

Advcd Theory and Sims

View full text Add to dashboard Cite

Technology computer‐aided design (TCAD) simulation is an important tool for the development of semiconductor devices. Based on coupled partial differential equations (PDEs) for behaviors, TCAD can calculate objects such as impurities, point defects, and electronic carriers in semiconductors. However, over recent years semiconductor devices have become increasingly miniaturized and complicated, resulting in much longer calculation times for TCAD. Machine learning is one technology that may be used to overcome this simulation cost problem. In this study, a neural network architecture is proposed that considers the structure of the coupled PDEs. Features representing each concentration distribution of the calculation objects are extracted by convolution operations and their reaction is modeled by channel attention. The performance of the proposed architecture and of conventional neural network models is evaluated using a simulation dataset generated by 1D coupled PDEs that models the diffusion and reaction of vacancies and interstitial atoms. In addition, the advantage of the method is discussed through the analyses of error correlations of the two predictions and attention coefficients. The machine learning method developed in this study will be applicable to other physics described by coupled PDEs and is expected to speed up the computation of simulations in various fields.

show abstract

Restructuring TCAD System: Teaching Traditional TCAD New Tricks

Cited by 9 publications

References 11 publications

Comprehensive studies on deep learning applicable to TCAD

Comprehensive studies on deep learning applicable to TCAD

PAC-Net: A Model Pruning Approach to Inductive Transfer Learning

Machine Learning for Semiconductor Process Simulation Described by Coupled Partial Differential Equations

Contact Info

Product

Resources

About