Computing device for predicting a profile using deep learning and operating method thereof

Patent metadata

Field	Value
Publication number	US-12299869-B2
Application number	US-202217849617-A
Country	US
Kind code	B2
Filing date	Jun 25, 2022
Priority date	Dec 17, 2021
Publication date	May 13, 2025
Grant date	May 13, 2025

Abstract

An operating method of a computing device for predicting a profile using deep learning includes sampling a unique pattern in a full chip, extracting a contour of a resist profile of each of a plurality of heights by performing rigorous simulation corresponding to the unique pattern, preparing an input image and an output image corresponding to the contour of each of the plurality of heights, performing deep learning on the extracted contour using the input image and the output image, and generating a profile prediction model according to performing of the deep leaning.

First claim

What is claimed is:

1. An operating method of a computing device for predicting a profile using deep learning, the operating method comprising:
sampling coordinates of a unique pattern across a full chip;
extracting a contour of a resist profile of each of a plurality of heights by performing rigorous simulation at each coordinate of the unique pattern;
preparing an input image and an output image corresponding to the contour of each of the plurality of heights;
performing deep learning on the extracted contour using the input image and the output image; and
generating a profile prediction model according to the performing of the deep learning,
wherein the sampled coordinates are less than all coordinates of the full chip.

2. The method of claim 1 , further comprising storing the extracted contour as a file in a type of a graphics design system (GDS).

3. The method of claim 1 ,
wherein the input image includes a single channel image, and
wherein the output image includes a multi-channel image.

4. The method of claim 1 , wherein the input image is used in an optical proximity corrected mask or a target database.

5. The method of claim 1 , wherein the output image is used in a contour for each height generated through the rigorous simulation.

6. The method of claim 1 , wherein the performing the deep learning includes performing model learning using a deep convolutional generative adversarial network (DCGAN).

7. The method of claim 1 , wherein the profile prediction model includes a multi-channel image-based deep learning model.

8. The method of claim 1 , further comprising evaluating a resist 3D profile for the full chip using the profile prediction model.

9. The method of claim 1 , further comprising verifying a defect by predicting a contour of a photo resist for each height.

10. The method of claim 1 , further comprising performing an etch model simulation by predicting a contour of a photo resist for each height.

11. A computing device for predicting a profile using deep learning, the computing device comprising:
at least one processor configured to execute a photo resist three-dimensional (3D) profile modeling tool; and
a memory configured to store the photo resist 3D profile modeling tool,
wherein when the at least one processor executes the photo resist 3D profile modeling tool, the at least one processor:
samples coordinates of a unique pattern across a full chip;
extracts a contour of a resist profile of each of a plurality of heights by performing a rigorous simulation at each coordinate of the unique pattern;
prepares an input image and an output image corresponding to the contour of each of the plurality of heights;
performs deep learning on the extracted contour using the input image and the output image; and
generates a profile prediction model according to performing of the deep learning,
wherein the sampled coordinates are less than all coordinates of the full chip.

12. The computing device of claim 11 , wherein the profile prediction model includes a multi-channel image-based deep learning model.

13. The computing device of claim 11 , wherein the input image is a single channel image, and the output image is a 3-channel image.

14. The computing device of claim 11 , wherein an optical proximity corrected prediction model is verified using the profile prediction model.

15. The computing device of claim 11 , wherein a hot spot defect is detected using the profile prediction model generated for the full chip.

16. An operating method of a computing device for predicting a profile using deep learning, the operating method comprising:
measuring sample data;
generating a contour for each height by performing rigorous simulation on a layout corresponding to the sample data;
performing multi-channel image-based deep learning model learning on the contour for each height; and
predicting a tilt of a resist profile using the contour for each height through the multi-channel image-based deep learning model learning.

17. The operating method of claim 16 , further comprising predicting a height at which an overlap is formed through contour prediction based on the height.

18. The operating method of claim 17 , further comprising determining a hotspot in consideration of a degree of the overlap and a margin of a pattern.

19. The operating method of claim 16 , further comprising generating a resist 3D profile prediction model through the multi-channel image-based deep learning model learning.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Lee Hyeok
• Yang Jaewon
• Yeo Sangchul
• Kim Eunju
• Lee Sooryong

Classifications

• G06V10/10

  Image acquisition (document image scanning and transmission H04N1/00; control of digital cameras H04N23/60) · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06F30/27

  using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model · CPC title

• G06T2207/30148

  Semiconductor; IC; Wafer · CPC title

• G06V10/46 Primary

  Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features (colour feature extraction G06V10/56) · CPC title