Unlock instant, AI-driven research and patent intelligence for your innovation.

Modeling method of chemical mechanical polishing chip surface height prediction model based on transfer learning

A chemical-mechanical, surface-height technology, applied in neural learning methods, biological neural network models, CAD circuit design, etc., can solve the time-consuming and labor-intensive problems of testing chips, and achieve faster convergence, improved speed and accuracy, and improved prediction The effect of precision

Pending Publication Date: 2021-12-14
SHANGHAI JIAO TONG UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since deep learning is a data-driven process, building an accurate and robust CMP chip surface height prediction model requires a large amount of experimental data, and usually requires the design of a special test chip to obtain experimental data. Time-consuming and labor-intensive test chips with similar layout
In addition, the layout characteristics of different circuit types, it is difficult for a single model to obtain accurate prediction results in different circuit types

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Modeling method of chemical mechanical polishing chip surface height prediction model based on transfer learning
  • Modeling method of chemical mechanical polishing chip surface height prediction model based on transfer learning
  • Modeling method of chemical mechanical polishing chip surface height prediction model based on transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] This embodiment implements a method for modeling the surface height prediction model of chemical mechanical polishing chips with different process parameters based on transfer learning.

[0043] figure 1 It is a schematic diagram of the transfer learning process of a method for modeling the surface height prediction model of chemical mechanical polishing chips with different process parameters based on transfer learning. as attached figure 1 As shown, in this embodiment, a method for modeling the surface height prediction model of a chemical mechanical polishing chip with different process parameters based on transfer learning, the transfer learning process specifically includes the following steps:

[0044] S1. Data collection: prepare a test chip, and collect simulation data of the test chip after chemical mechanical polishing.

[0045] In this embodiment, the test chip adopts a mixed-signal commercial chip including analog circuits, digital circuits and memory circ...

Embodiment 2

[0058] This embodiment implements a method for modeling the surface height prediction model of a specific circuit type chemical mechanical polishing chip based on transfer learning.

[0059] T1. In this embodiment, a specific circuit type chemical mechanical polishing chip surface height prediction model modeling method based on transfer learning is based on steps S1-S3 of Embodiment 1, using a mixed-signal commercial chip containing analog circuits, digital circuits and storage circuits as The test chip establishes the source domain data set, performs model training and hyperparameter adjustment through the training set and verification set, and obtains the surface height prediction model of the chemical mechanical polishing chip in the source domain, and uses the test set to evaluate the prediction effect of the trained model.

[0060] T2. Establish target domain data sets. According to steps S1-S2 in Embodiment 1, test chips of different circuit types are used to respectivel...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a method for modeling a surface height prediction model of a chemical mechanical polishing chip based on transfer learning. The method comprises the following steps: collecting chemical mechanical polishing simulation data, preprocessing the data, and establishing a source domain data set; performing model training based on the source domain data set by using the designed neural network model, and establishing a source domain chemical mechanical polishing chip surface height prediction model; establishing a target domain data set based on different chemical mechanical polishing process parameters or different types of circuits, and generating a target domain chemical mechanical polishing chip surface height prediction model by using a transfer learning method; the beneficial effects are that the circuit adapts to technological parameter changes and circuit type differences, and the versatility is high.

Description

【Technical field】 [0001] The invention relates to the technical field of semiconductor chemical mechanical polishing technology, in particular to a modeling method for predicting the surface height of a chemical mechanical polishing chip based on migration learning. 【Background technique】 [0002] As the integrated circuit semiconductor manufacturing process enters the stage of deep submicron (usually 0.35-0.8μm and below are called submicron level, and 0.25um and below are called deep submicron), the feature size of the circuit is further reduced, and manufacturing The degree of process deviation in the process is becoming more and more serious, which greatly affects the performance and yield of chips. Among them, chemical mechanical polishing (CMP), as a key technology to achieve a high degree of planarization of the chip surface, mainly realizes the planarization of the chip surface topography through the mechanical force between the polishing pad and the silicon wafer an...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/38G06N3/04G06N3/08
CPCG06F30/38G06N3/08G06N3/045
Inventor 李永福张晴黄华杰王国兴连勇
Owner SHANGHAI JIAO TONG UNIV