CN101118422A - Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system - Google Patents
Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system Download PDFInfo
- Publication number
- CN101118422A CN101118422A CNA2006101084086A CN200610108408A CN101118422A CN 101118422 A CN101118422 A CN 101118422A CN A2006101084086 A CNA2006101084086 A CN A2006101084086A CN 200610108408 A CN200610108408 A CN 200610108408A CN 101118422 A CN101118422 A CN 101118422A
- Authority
- CN
- China
- Prior art keywords
- data
- quality control
- fault detection
- classification
- virtual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 92
- 239000004065 semiconductor Substances 0.000 title description 6
- 238000003908 quality control method Methods 0.000 claims abstract description 50
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000005070 sampling Methods 0.000 claims abstract description 17
- 238000001514 detection method Methods 0.000 claims description 42
- 235000012431 wafers Nutrition 0.000 claims description 17
- 238000010586 diagram Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 2
- 238000009966 trimming Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 10
- 238000005259 measurement Methods 0.000 description 7
- 238000003070 Statistical process control Methods 0.000 description 6
- 238000002872 Statistical quality control Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Landscapes
- General Factory Administration (AREA)
Abstract
The present invention discloses a method for creating a pre-estimation model. The method is to choose the varying detecting value of a plurality of real time technique states, and collect error detecting and categorical data as well as quality control data according to different techniques. The data control quality is judged whether the data control quality measures up to standard. When the data control quality does not measure up to standard, the error detecting and categorical data and the quality control data are recollected, or the sampling frequency is trimmed. When the data control quality measures up to standard, the error detecting and categorical data and the quality control data are processed. A plurality of pre-estimation models is created according to the data processing result, an optimum pre-estimation model is chosen according to a pertinence indicating hand to validate the pre-estimation model, and optimum pre-estimation model is trimmed. When the trimming is finished and the optimum pre-estimation model is in the optimum state, the optimum pre-estimation model is delivered to a dummy measure engine.
Description
Technical Field
The present invention relates to a quality control method for semiconductor manufacturing, and more particularly, to a method and system for virtual metrology estimation and estimation modeling for semiconductor manufacturing.
Background
Statistical Quality Control (SQC) is a technology for maintaining and improving product Quality, and Statistical Process Control (SPC) is one of the main tools, which focuses on the analysis of data during the manufacturing Process to determine the cause of product variation. Statistical quality management consists of two major components, statistical process control and sampling admission criteria. Statistical process control includes Quality Control (QC) processing and basic theories of probability and statistics and their applications. SPC is used to predictively monitor production variables or quality variables of products using process operating variables, and since there is a certain time lag between the time when the process operating variables change and the time when safety/quality issues arise, how to predict the quality variables in the shortest time is one of the important factors to be considered when evaluating the relative merits of SPC methods.
In addition, in the production process, a plurality of differences are allowed, only the differences need to be properly controlled, and the quality degree needs to be controlled so as to reach a certain required range. The Quality Control (QC) is to extract samples during the manufacturing process, to perform statistical analysis on the data obtained from the sample measurement and to draw a control chart to control whether the process is abnormal or not, or to extract several samples from a large batch of products and to check the characteristics of the samples, and to analyze the obtained data to determine whether the products are qualified or not and whether the products need to be disposed.
Conventional metrology methods can lead to a number of application-related problems due to sampling differences, such as lot control between two wafer lots. Therefore, the present invention provides a method and system for estimating and establishing an estimation model for virtual metrology in semiconductor manufacturing, which can reduce Sampling Risk (Sampling Risk) caused by Sampling and reduce Sampling frequency (Sampling Rate) of a measurement station.
Disclosure of Invention
Based on the above objective, the embodiment of the present invention discloses a method for building a prediction model. Selecting a plurality of real-time process state variation detection values of a machine according to different processes, collecting the selected real-time process state variation detection values into error detection and classification data, and collecting quality control data. And judging whether the data control quality meets the standard or not according to the collected fault detection and classification data and the quality control data. If not, re-collecting the fault detection and classification data and quality control data or fine-tuning the sampling frequency. If the data meets the standard, the fault detection and classification data and the quality control data are processed. Establishing a plurality of estimation models according to the data processing result, selecting an optimal estimation model according to a correlation pointer, verifying the estimation model, and finely adjusting the optimal estimation model. If the fine tuning is completed and the optimal pre-estimated model is in the optimal state, the optimal pre-estimated model is transmitted to a virtual measurement engine.
The embodiment of the invention also discloses a method for estimating the quality control data by the virtual measurement engine. Error detection and classification data are collected, and whether the data control quality meets the standard is judged according to the error detection and classification data. If not, stopping the estimation process. If the data meets the standard, the collected fault detection and classification data is processed, and the estimation of the virtual quality control is executed according to the virtual quality control data obtained after the data processing.
The embodiment of the invention also discloses a virtual engine pre-estimation system, which comprises a first process machine, a second process machine, a fault detection and classification system, a virtual measurement engine and a quality control measurement station. The first process machine executes required processes on a batch of input wafers. The fault detection and classification system collects fault detection and classification data and quality control data of the batch of wafers from the first process tool. The virtual measurement engine obtains the collected fault detection and classification data and quality control data from the fault detection and classification system to predict the batch of wafers, and if the prediction result is normal, the batch of wafers is output to the second process machine to execute another process. If the estimated result is not normal, the quality control measurement station samples the fault detection and classification data and the quality control data to determine whether the collected fault detection and classification data and quality control data are available, if not, the fault detection and classification data and the quality control data of the batch of chips are collected again, and if so, the batch of chips are output to the second process machine to execute another process.
Drawings
Fig. 1 shows a real-time process state variation detection value of a tool.
FIG. 2 is a schematic diagram of a virtual metrology engine according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating steps performed in building a predictive model according to an embodiment of the invention.
FIG. 4 is a flow chart showing steps of data processing according to an embodiment of the present invention.
FIG. 5 is a flowchart illustrating an implementation procedure of the virtual metrology engine estimating QC data according to an embodiment of the present invention.
Fig. 6 is a schematic diagram illustrating an architecture of a virtual engine forecast system according to an embodiment of the present invention.
Description of the figures
600-virtual engine prediction system
610-first process machine
620-error detection and classification system
630-virtual measurement Engine
640-quality control measurement station
650-second process machine
660 to other hosts
Detailed Description
In order to make the objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with fig. 1 to 6 are described in detail below. The present description provides different examples to illustrate the technical features of different embodiments of the present invention. The arrangement of the components in the embodiments is for illustration and not for limiting the invention. And the reference numbers in the embodiments are partially repeated to simplify the description, and do not indicate the relevance between the different embodiments.
The embodiment of the invention discloses a method and a system for virtual measurement prediction and prediction model establishment in semiconductor manufacturing.
To obtain detailed process tool performance data, a large number of high resolution system variables, i.e., a real-time process state Identification (SVID) value for each tool, such as temperature, pressure, gas flow, etc., are required, as shown in fig. 1. An error Detection and Classification (FDC) system is used to collect the SVID of each tool (i.e. error Detection and Classification Data (FDC Data)), and then the obtained FDC Data is transmitted to a Virtual measurement Engine (Virtual measurement Engine) of the present invention, and a Virtual measurement result is generated after calculation, as shown in fig. 2.
Virtual Metrology (VM) is a method of applying intelligent computing (computerized Intelligence) to estimate the quality of a project processed by a manufacturing tool in real time. The accuracy of the estimation can be based on the general advanced process and Equipment control technology (advanced process/Equipment)Control, APC/AEC), the virtual metrology can replace conventional metrology tools. Therefore, the virtual measurement engine of the present invention is a system with a built-in estimation model, and when the FDC data is input into the virtual measurement engine, a virtual measurement result can be generated. The predictive model includes(which is used to estimate film Thickness (Thickness)), CD = f (x) 1 ，x 2 ，…，x n ADI) (which is used to estimate the line width (CD)), and the like, wherein x i Denoted as SVID and ADI is post-development process chip inspection (after development inspection). The following describes an implementation process of how to build the predictive model.
FIG. 3 is a flowchart illustrating steps performed in building predictive models according to an embodiment of the present invention.
First, the SVID of the tool, for example, temperature, pressure, gas flow rate, etc., is selected according to the different processes (step S11). The selected SVIDs are then collected into FDC data using the FDC system, and Quality Control (QC) data, such as film thickness, line width, etc., are collected (step S12). Next, it is judged whether the quality of the data control product meets the standard or not based on the collected FDC data and QC data (step S13). If not, go back to step S12 to collect FDC data and QC data or fine tune the sampling frequency. Non-compliance means less than 95% of the standard. If the FDC data and QC data are met, the collected FDC data and QC data are processed (step S14). The data processing comprises the following steps. As shown in FIG. 4, first, the original process data and the control chart thereof are inspected (step S21). Among all the parameters or factors that may affect the process, a parameter or factor having a large degree of influence on the process is selected, and a corresponding function is selected in a predefined function database (step S22). Next, data processing is performed according to the selected parameter or factor and the corresponding function (step S23).
Next, a predictive model is built based on the data processing result (step S15). The method for establishing the prediction model also comprises a Multiple Linear Regression (MLR) method and a Partial Least Squares (PLS) method.
The complex regression formula is expressed as follows:
wherein x i Denoted as SVID. A is available in the data processing step (S14) i Then a is added to i And substituting the SVID and the SVID into the complex regression formula to obtain a predicted value.
The partial least squares formula is expressed as follows:
Prediction＝f(x 1 ，x 2 ，…，x n ADI) in which x i Denoted as SVID. Similarly, by substituting SVI and ADI into the above-mentioned partial least squares formula, the predicted value can be obtained. The partial least squares method can predict multiple data simultaneously at one time point, and can process raw data (RawData) and real-time data (SummaryData).
If the predictive model fails to be established (NG), and an incorrect SVID may be collected, the process must return to step S11 to reselect the SVID.
After the pre-estimation model is built, a best pre-estimation model is selected according to a correlation index (R-Square) (step S16). When the R-Square of a prediction model is less than 0.7, which indicates that the prediction model is a bad model and cannot provide a correct prediction result (NG), the method must return to step S11 to reselect the SVID. After a best prediction model is selected, the best prediction model is verified (step S17). When the verification is completed, the best predictive model is fine-tuned (step S18). The best predictive model is trimmed according to the load scatter Plot (Loading Plot), the Contribution scatter Plot (Contribution Plot) or the Single-squared partial PLS (Single Variable PLS), and if the best predictive model is not in the best state (NG, the first time), which may be a problem in the data processing of the previous step S14, after trimming, the data processing is performed again in step S14. If the estimated model is still not in the best state (NG, second time) after the second fine tuning, go back to step S11 to re-execute the above process. If the fine tuning is completed and the best pre-estimated model is in the best state, the best pre-estimated model is transmitted to the virtual measurement engine (step S19).
The process of the virtual metrology engine predicting the QC data is described next.
As shown in fig. 5, after the virtual metrology engine is built, it is first necessary to collect FDC data (step S31), and then determine whether the data control quality meets the standard (step S32). If the standard value is not met, i.e. less than 95% of the standard value, the estimation process is stopped (step S33). If the FDC data meets the standard, the collected FDC data is processed (step S34), the estimation of the virtual QC is executed according to the virtual quality control data obtained after the data processing is finished (step S35), and then the virtual QC data is transmitted to other hosts to be used as other applications (step S36).
Fig. 6 is a schematic diagram illustrating an architecture of a virtual engine forecast system according to an embodiment of the present invention.
The virtual engine estimation system 600 of the embodiment of the invention includes a first process machine 610, an FDC system 620, a virtual measurement engine 630, a QC measurement station 640, and a second process machine 650. A lot of wafers is first transferred to the first processing tool 610 to perform a desired process, and the FDC system 620 collects FDC data and QC data of the lot of wafers from the first processing tool 610 and then transfers the collected FDC data and QC data to the virtual metrology engine 630 for pre-evaluation of the lot of wafers. If the estimated result is normal, the batch of wafers may be output to the second process tool 650 for performing another process. If the estimation result is not normal, the QC measurement station 640 samples the FDC data and the QC data to determine whether the collected FDC data and QC data are available. If not, the FDC data and QC data for the batch of wafers are re-collected. If available, the batch of wafers may be output to the second process tool 650 for another process. The virtual metrology engine 630 may also send 660 the generated virtual QC data to other hosts for other applications.
The method and the system for virtually measuring, estimating and establishing the estimation model can reduce Sampling Risk (Sampling Risk) caused by Sampling and can simultaneously reduce the Sampling frequency (Sampling Rate) of a measuring station.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.
Claims (11)
1. A method for establishing a prediction model comprises the following steps:
selecting a plurality of real-time process state variation detection values of a machine according to different processes;
collecting the real-time process state variation detection values to form fault detection and classification data, and collecting quality control data;
judging whether the data control quality meets the standard according to the collected fault detection and classification data and the quality control data;
if not, re-collecting the error detection and classification data and quality control data or fine-tuning the sampling frequency;
if the error detection and classification data meets the standard, processing the error detection and classification data and the quality control data;
establishing a plurality of pre-estimation models according to the data processing result;
selecting an optimal pre-estimation model according to a correlation pointer; and
the best pre-estimated model is sent to a virtual metrology engine.
2. The method of claim 1, further comprising validating the predictive model.
3. The method of building predictive models of claim 2, further comprising the steps of:
fine-tuning the optimal pre-estimation model; and
if the fine tuning is completed and the best pre-estimated model is in the best state, the best pre-estimated model is transmitted to the virtual measurement engine.
4. A method of building predictive models as claimed in claim 3, further comprising the steps of:
fine-tuning the optimal pre-prediction model according to a load scatter diagram, a contribution scatter diagram or a partial least square single variable;
after fine tuning, if the optimal pre-estimated model is not in the optimal state, re-executing data processing, and then fine tuning again; and
after fine tuning, if the best pre-estimated model is not in the best state, the flow of establishing the pre-estimated model is executed again.
5. The method for building predictive models of claim 1, wherein the data processing further comprises the steps of:
inspecting original process data and a control graph thereof;
selecting parameters or factors which have a greater influence on the process from all parameters or factors which can influence the process, and selecting corresponding function functions from a predefined function database; and
and processing data according to the selected parameters or factors and the corresponding function functions.
6. The method of claim 1, wherein the predictive model is built using a multiple regression method and a partial least squares method.
7. The method of creating a predictive model of claim 6 wherein the partial least squares method is used to predict multiple data simultaneously at a time point and process both raw and real time data.
8. The method of claim 1, wherein if a failure of the predictive model is detected, the real-time process state variation detection value is reselected.
9. A method for predicting quality control data by a virtual measurement engine comprises the following steps:
collecting fault detection and classification data;
judging whether the data control quality meets the standard according to the fault detection and the classification data;
if not, stopping the estimation flow;
if the standard is met, performing data processing on the collected fault detection and classification data; and
and performing virtual quality control pre-estimation according to the virtual quality control data obtained after data processing.
10. The method of claim 9, further comprising sending the virtual quality control data to other hosts for other applications.
11. A virtual engine projection system, comprising:
a first process machine for executing required process to a batch of input wafers;
a second process machine;
a fault detection and classification system, coupled to the first process tool, for collecting fault detection and classification data and quality control data of the batch of wafers from the first process tool;
a virtual measurement engine, coupled to the fault detection and classification system, for obtaining collected fault detection and classification data and quality control data from the fault detection and classification system to predict the batch of wafers, and outputting the batch of wafers to the second process machine to execute another process if the prediction result is normal;
and a quality control measuring station coupled to the virtual measuring engine for sampling the fault detection and classification data and quality control data to determine whether the collected fault detection and classification data and quality control data are available if the estimated result is abnormal, re-collecting the fault detection and classification data and quality control data of the batch of wafers if the collected fault detection and classification data and quality control data are unavailable, and outputting the batch of wafers to the second process machine if the collected fault detection and classification data and quality control data are available to execute another process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101084086A CN101118422A (en) | 2006-08-02 | 2006-08-02 | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101084086A CN101118422A (en) | 2006-08-02 | 2006-08-02 | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101118422A true CN101118422A (en) | 2008-02-06 |
Family
ID=39054567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006101084086A Pending CN101118422A (en) | 2006-08-02 | 2006-08-02 | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101118422A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102063063A (en) * | 2009-11-11 | 2011-05-18 | 台湾积体电路制造股份有限公司 | Semiconductor manufacturing method and system |
| CN101853776B (en) * | 2009-03-31 | 2011-12-28 | 台湾积体电路制造股份有限公司 | Advanced process control with novel sampling policy |
| CN102456083A (en) * | 2010-10-20 | 2012-05-16 | 北京北方微电子基地设备工艺研究中心有限责任公司 | Process data analyzing method and system |
| CN103278714A (en) * | 2013-05-15 | 2013-09-04 | 江苏大学 | Virtual measurement method and system for mixed manufacturing process |
| CN103592913A (en) * | 2013-10-30 | 2014-02-19 | 江苏大学 | Machine bench performance matching method and system of semiconductor manufacturing equipment |
| CN104952750A (en) * | 2014-03-26 | 2015-09-30 | 中芯国际集成电路制造(上海)有限公司 | Early-stage detecting system and method for silicon chip electrical test |
| CN107644823A (en) * | 2016-07-21 | 2018-01-30 | 株式会社日立国际电气 | The manufacture method of lining processor and semiconductor devices |
| CN108630238A (en) * | 2017-03-22 | 2018-10-09 | 株式会社东芝 | Manufacturing method, multilayer film film-forming system and the film forming method of adjustment of magnetic recording media |
| CN109444703A (en) * | 2018-10-15 | 2019-03-08 | 上海华虹宏力半导体制造有限公司 | The test method of super-junction device |
| CN113094893A (en) * | 2021-04-06 | 2021-07-09 | 江苏大学 | Wafer quality virtual measurement method and device, computer equipment and storage medium |
-
2006
- 2006-08-02 CN CNA2006101084086A patent/CN101118422A/en active Pending
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853776B (en) * | 2009-03-31 | 2011-12-28 | 台湾积体电路制造股份有限公司 | Advanced process control with novel sampling policy |
| CN102063063A (en) * | 2009-11-11 | 2011-05-18 | 台湾积体电路制造股份有限公司 | Semiconductor manufacturing method and system |
| CN102063063B (en) * | 2009-11-11 | 2016-07-06 | 台湾积体电路制造股份有限公司 | Semiconductor making method and system |
| CN102456083A (en) * | 2010-10-20 | 2012-05-16 | 北京北方微电子基地设备工艺研究中心有限责任公司 | Process data analyzing method and system |
| CN102456083B (en) * | 2010-10-20 | 2013-10-30 | 北京北方微电子基地设备工艺研究中心有限责任公司 | Process data analyzing method and system |
| CN103278714B (en) * | 2013-05-15 | 2015-10-28 | 江苏大学 | A kind of virtual measurement method and system mixing processing procedure |
| CN103278714A (en) * | 2013-05-15 | 2013-09-04 | 江苏大学 | Virtual measurement method and system for mixed manufacturing process |
| CN103592913B (en) * | 2013-10-30 | 2015-10-28 | 江苏大学 | The board Performance Match method and system of semiconductor manufacturing facility |
| CN103592913A (en) * | 2013-10-30 | 2014-02-19 | 江苏大学 | Machine bench performance matching method and system of semiconductor manufacturing equipment |
| CN104952750A (en) * | 2014-03-26 | 2015-09-30 | 中芯国际集成电路制造(上海)有限公司 | Early-stage detecting system and method for silicon chip electrical test |
| CN104952750B (en) * | 2014-03-26 | 2017-11-24 | 中芯国际集成电路制造(上海)有限公司 | The early stage detecting system and method for a kind of silicon chip electrical testing |
| CN107644823A (en) * | 2016-07-21 | 2018-01-30 | 株式会社日立国际电气 | The manufacture method of lining processor and semiconductor devices |
| CN107644823B (en) * | 2016-07-21 | 2021-01-26 | 株式会社国际电气 | Substrate processing apparatus and method for manufacturing semiconductor device |
| CN108630238A (en) * | 2017-03-22 | 2018-10-09 | 株式会社东芝 | Manufacturing method, multilayer film film-forming system and the film forming method of adjustment of magnetic recording media |
| US10907248B2 (en) | 2017-03-22 | 2021-02-02 | Kabushiki Kaisha Toshiba | Manufacturing method of magnetic recording medium, film formation system of multilayer film, and film formation adjustment method |
| CN109444703A (en) * | 2018-10-15 | 2019-03-08 | 上海华虹宏力半导体制造有限公司 | The test method of super-junction device |
| CN113094893A (en) * | 2021-04-06 | 2021-07-09 | 江苏大学 | Wafer quality virtual measurement method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101118422A (en) | Virtual measurement prediction generated by semi-conductor, method for establishing prediction model and system | |
| US8452441B2 (en) | Process quality predicting system and method thereof | |
| CN112382582B (en) | Wafer test classification method and system | |
| TWI549007B (en) | Method for searching and analyzing process parameters and computer program product thereof | |
| US8437870B2 (en) | System and method for implementing a virtual metrology advanced process control platform | |
| TWI443776B (en) | An automated state estimation system for cluster tools and a method of operating the same | |
| KR20050085716A (en) | Dynamic adaptive sampling rate for model prediction | |
| CN110503288B (en) | System and method for identifying yield loss reason considering machine interaction | |
| KR102003961B1 (en) | System and method for identifying root causes of yield loss | |
| JP4568786B2 (en) | Factor analysis apparatus and factor analysis method | |
| CN105702595B (en) | The yield judgment method of wafer and the changeable quantity measuring method of wafer conformity testing | |
| KR20090001148A (en) | Virtual metrology system and virtual metrology method | |
| Pan et al. | A virtual metrology system for predicting end-of-line electrical properties using a MANCOVA model with tools clustering | |
| JP2020061575A (en) | System for determining critical parameters using high-dimensional variable selection model | |
| KR100455559B1 (en) | Method and system for in-line monitoring process performance using measurable equipment signals | |
| KR101998972B1 (en) | Method of analyzing and visualizing the cause of process failure by deriving the defect occurrence index by variable sections | |
| CN115032493A (en) | Wafer testing method and system based on tube core parameter display | |
| CN113742248A (en) | Method and system for predicting organization process based on project measurement data | |
| TWI427487B (en) | Method for sampling workpiece for inspection and computer program product performing the same | |
| CN111539521A (en) | Method for predicting yield of semiconductor product by neural network error-back propagation algorithm | |
| CN103278714B (en) | A kind of virtual measurement method and system mixing processing procedure | |
| CN101118423A (en) | Serviceability selecting method and system of virtual measuring prediction model | |
| Kim et al. | AI-guided reliability diagnosis for 5, 7nm automotive process | |
| Lee et al. | Reliability Prediction for Automotive 5nm and 7nm Technology node by using Machine Learning based Solution | |
| TWI647770B (en) | Yield rate determination method for wafer and method for multiple variable detection of wafer acceptance test |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |