CN117094095A - Wafer warpage optimization method and system - Google Patents
Wafer warpage optimization method and system Download PDFInfo
- Publication number
- CN117094095A CN117094095A CN202311342862.8A CN202311342862A CN117094095A CN 117094095 A CN117094095 A CN 117094095A CN 202311342862 A CN202311342862 A CN 202311342862A CN 117094095 A CN117094095 A CN 117094095A
- Authority
- CN
- China
- Prior art keywords
- wafer
- warpage
- optimization
- information
- production
- 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.)
- Granted
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 134
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000004519 manufacturing process Methods 0.000 claims abstract description 152
- 238000009826 distribution Methods 0.000 claims abstract description 48
- 238000004458 analytical method Methods 0.000 claims abstract description 30
- 230000010354 integration Effects 0.000 claims abstract description 8
- 238000013507 mapping Methods 0.000 claims description 17
- 238000002372 labelling Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 13
- 238000005259 measurement Methods 0.000 claims description 12
- 238000010586 diagram Methods 0.000 claims description 11
- 230000000877 morphologic effect Effects 0.000 claims description 9
- 238000005452 bending Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 5
- 230000004927 fusion Effects 0.000 claims description 4
- 238000003709 image segmentation Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 13
- 230000000694 effects Effects 0.000 abstract description 2
- 235000012431 wafers Nutrition 0.000 description 325
- 239000010410 layer Substances 0.000 description 14
- 239000004065 semiconductor Substances 0.000 description 6
- 238000013433 optimization analysis Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000011217 control strategy Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 238000004886 process control Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 238000000137 annealing Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000000919 ceramic Substances 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000002346 layers by function Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000011165 process development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/18—Manufacturability analysis or optimisation for manufacturability
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Analysis (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computational Mathematics (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
Abstract
The application relates to the technical field of data processing, and provides a wafer warpage optimization method and system. The method comprises the following steps: carrying out loss analysis on the wafer warpage distribution map and the standard wafer warpage distribution map based on the digital twin network model to obtain wafer warpage loss characteristic information; based on the wafer warpage loss characteristic information, generating wafer warpage process influence factor information, and carrying out data searching integration to construct a wafer production strategy database; performing similarity matching according to the wafer attribute characteristic information and a wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space; and performing global optimization in a wafer warpage optimization space based on the warpage optimization fitness function to generate wafer warpage optimization parameter information so as to perform wafer optimization production control. By adopting the method, the technical effects of improving the accuracy and comprehensiveness of wafer warpage optimization parameter analysis and further effectively and reliably controlling and adjusting the wafer warpage can be achieved.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a wafer warpage optimization method and a wafer warpage optimization system based on the attribute characteristic information system.
Background
Wafer warpage refers to the degree of bending of a wafer during processing or use, and various stresses are generated on the surface of the wafer along with the processing of a semiconductor, so that the wafer warps during the processing, which is an important parameter in the semiconductor manufacturing process. The wafer is used as a substrate of the chip, and the flatness warping degree of the wafer can influence whether the chip technology is completely displayed or not, so that the flatness warping degree of the wafer directly influences the quality of the chip. However, the wafer warpage optimization analysis in the prior art has low accuracy, so that the wafer warpage cannot be effectively controlled.
Disclosure of Invention
Based on the above, it is necessary to provide a wafer warpage optimization method and system capable of realizing warpage parameter optimization analysis by using wafer warpage loss characteristics, improving accuracy and comprehensiveness of the optimization parameter analysis, and further effectively and reliably controlling and adjusting wafer warpage.
A method of wafer warp optimization, the method comprising: obtaining a warp distribution map of a target wafer through wafer warp stress measurement equipment; obtaining a standard wafer warpage distribution map according to a wafer application standard, and carrying out loss analysis on the warpage distribution map and the standard wafer warpage distribution map based on a digital twin network model to obtain wafer warpage loss characteristic information; analyzing the production process flow of the target wafer based on the wafer warpage loss characteristic information to generate wafer warpage process influence factor information; performing data searching integration based on the wafer warpage process influence factor information to construct a wafer production strategy database; performing similarity matching according to the attribute characteristic information of the target wafer and the wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space; and constructing a warpage optimization fitness function, performing global optimization in the wafer warpage optimization space based on the warpage optimization fitness function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information.
A wafer warp optimization system, the system comprising: the warp distribution map acquisition module is used for acquiring a warp distribution map of the target wafer through wafer warp stress measurement equipment; the warpage loss analysis module is used for acquiring a standard wafer warpage distribution diagram according to a wafer application standard, and carrying out loss analysis on the warpage distribution diagram and the standard wafer warpage distribution diagram based on a digital twin network model to acquire wafer warpage loss characteristic information; the process influence factor generation module is used for analyzing the production process flow of the target wafer based on the wafer warpage loss characteristic information to generate wafer warpage process influence factor information; the strategy database construction module is used for carrying out data search integration based on the wafer warpage process influence factor information to construct a wafer production strategy database; the warpage optimization space obtaining module is used for carrying out similarity matching according to the attribute characteristic information of the target wafer and the wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space; and the wafer optimization production control module is used for constructing a warpage optimization adaptability function, performing global optimization in the wafer warpage optimization space based on the warpage optimization adaptability function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information.
The wafer warpage optimization method and the wafer warpage optimization system solve the technical problem that the wafer warpage cannot be effectively controlled due to low accuracy of wafer warpage optimization analysis in the prior art, achieve the technical effects of realizing warpage parameter optimization analysis by utilizing wafer warpage loss characteristics, improving accuracy and comprehensiveness of optimized parameter analysis, and further effectively and reliably controlling and adjusting the wafer warpage.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
FIG. 1 is a flow chart of a wafer warpage optimization method in one embodiment;
FIG. 2 is a flow chart of obtaining wafer warpage loss characteristics in a wafer warpage optimization method in one embodiment;
FIG. 3 is a block diagram of a wafer warp optimization system in one embodiment;
reference numerals illustrate: the wafer optimizing production control system comprises a warp distribution map acquisition module 11, a warp loss analysis module 12, a process influence factor generation module 13, a strategy database construction module 14, a warp optimizing space acquisition module 15 and a wafer optimizing production control module 16.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
As shown in fig. 1, the present application provides a wafer warpage optimization method, which includes:
step S100: obtaining a warp distribution map of a target wafer through wafer warp stress measurement equipment;
specifically, the wafer warpage refers to the degree of bending of a wafer during a processing process or a use process, and various stresses are generated on the surface of the wafer along with the processing of a semiconductor, so that the wafer warps during the processing process, which is an important parameter in the semiconductor manufacturing process. The wafer is used as a substrate of the chip, and the flatness warping degree of the wafer can influence whether the chip technology is completely displayed or not, so that the flatness warping degree of the wafer directly influences the quality of the chip.
In order to realize accurate measurement of wafer warpage, wafer warpage stress measurement equipment is firstly obtained, and the wafer warpage stress measurement equipment has detection functions of three-dimensional warpage (flatness), film stress, nano-profile, macroscopic defect imaging and the like, is suitable for semiconductor wafer production, semiconductor process development and glass and ceramic wafer production, and can perform disposable non-contact full-caliber uniform sampling measurement on the surfaces of various wafers. And obtaining a warpage distribution map of the target wafer through multiple measurements of the wafer warpage stress measurement equipment, wherein the warpage distribution map is used for visually displaying surface warpage morphology distribution information of the wafer to be tested and providing a data basis for subsequent wafer warpage optimization.
Step S200: obtaining a standard wafer warpage distribution map according to a wafer application standard, and carrying out loss analysis on the warpage distribution map and the standard wafer warpage distribution map based on a digital twin network model to obtain wafer warpage loss characteristic information;
in one embodiment, as shown in fig. 2, the obtaining wafer warpage loss feature information, step S200 of the present application further includes:
step S210: building a digital twin network model, wherein the digital twin network model comprises a target twin network sub-model and a standard twin network sub-model;
step S220: inputting the warpage distribution map and the standard wafer warpage distribution map into the target twin network sub-model and the standard twin network sub-model respectively, and outputting a target wafer warpage feature set and a standard wafer warpage feature set;
step S230: acquiring a characteristic loss function, wherein the characteristic loss function is a characteristic comparison value weighted fusion function;
step S240: performing loss analysis on the target wafer warpage feature set and the standard wafer warpage feature set based on the feature loss function, and outputting wafer warpage loss data;
step S250: and classifying the wafer warpage loss data in a characteristic level to obtain the wafer warpage loss characteristic information.
In one embodiment, the outputting the target wafer warp feature set, step S220 of the present application further includes:
step S221: obtaining an image division logic layer and a feature extraction logic layer according to the target twin network sub-model;
step S222: determining a preset division fineness based on the image division logic layer, and performing image segmentation on the warpage profile according to the preset division fineness to obtain a grid division warpage profile;
step S223: determining preset convolution characteristic information of the wafer according to the wafer bending requirement factors;
step S224: performing traversal convolution calculation on the grid division warping distribution graph according to the wafer preset convolution characteristic information to obtain the target wafer warping characteristic set conforming to a preset convolution numerical range.
Specifically, a standard wafer warp profile is obtained according to a wafer application standard, wherein the standard wafer warp profile is warp morphology profile information conforming to the wafer warp application standard. And carrying out loss analysis on the warpage profile and the standard wafer warpage profile based on a digital twin network model, namely comparing and analyzing the produced wafer warpage profile. In order to realize quick feature comparison of the warping distribution diagram, a digital twin network model is firstly built, the digital twin network model is a network structure formed by splicing a neural network with the similarity of two samples and shared weight, and specifically comprises a target twin network sub-model and a standard twin network sub-model, the network structure of the target twin network sub-model is the same as that of the standard twin network sub-model, and the data training processing process is the same.
The warp distribution map and the standard wafer warp distribution map are respectively input into the target twin network sub-model and the standard twin network sub-model, and taking the processing process of the target twin network sub-model as an example, firstly, according to the target twin network sub-model, a model functional structure logic layer of the model functional structure logic layer is obtained, and the model functional structure logic layer comprises an image division logic layer and a feature extraction logic layer. The image dividing logic layer is used for carrying out grid division on the image, and the preset dividing fineness is determined based on the image dividing logic layer, wherein the preset dividing fineness is the number of image dividing grid lines, and the more the number of dividing grid lines is, the higher the corresponding image processing fineness is. And performing image segmentation on the warpage distribution map according to the preset division fineness to obtain the warpage distribution map after grid division, and improving the accuracy of image feature analysis.
And determining preset convolution characteristic information of the wafer according to the wafer bending requirement factors, namely taking the wafer bending requirement criteria as a convolution extraction characteristic of a distribution map. And performing traversal convolution calculation on the grid division warping distribution graph according to the preset convolution characteristic information of the wafer, wherein the preset convolution numerical range is that the image convolution characteristic accords with a judgment threshold value, and the preset convolution numerical range can be set by itself to obtain the image convolution characteristic which accords with the preset convolution numerical range as a target wafer warping characteristic set. The standard twin network sub-model and the target twin network sub-model have the same functional layer structure, and the output result of the standard twin network sub-model, namely the standard wafer warping feature set, is obtained according to the same processing mode.
And obtaining a characteristic loss function, wherein the characteristic loss function is a characteristic comparison value weighted fusion function, namely, carrying out average weighted calculation on the characteristic comparison difference value of the characteristic data of the two sample images, and taking the calculation result as the loss data of the wafer warpage characteristic set. And carrying out loss analysis on the target wafer warpage feature set and the standard wafer warpage feature set based on the feature loss function, and outputting wafer warpage loss data, namely, warpage feature data which does not accord with the wafer production standard. And classifying the wafer warpage loss data in a characteristic level, namely classifying the type and degree of warpage characteristics of the wafer warpage loss data, and determining the wafer warpage loss characteristic information. And the wafer warpage loss characteristics are rapidly analyzed by utilizing the digital twin network model, and the wafer warpage situation is accurately judged, so that the warpage optimization analysis accuracy is improved.
Step S300: analyzing the production process flow of the target wafer based on the wafer warpage loss characteristic information to generate wafer warpage process influence factor information;
in one embodiment, the generating wafer warpage process influencing factor information, step S300 of the present application further includes:
step S310: obtaining a wafer production warpage database, wherein the wafer production warpage database comprises a wafer warpage loss feature set and a wafer production process control parameter set;
step S320: performing simulation training based on the wafer warpage loss feature set and the wafer production process control parameter set to construct a wafer production warpage analyzer;
step S330: analyzing the wafer warpage loss characteristic information to obtain wafer warpage morphological characteristics and wafer warpage characteristics;
step S340: and analyzing the warpage influence of the wafer warpage morphological characteristics and the wafer warpage characteristics based on the wafer production warpage analyzer, and determining the wafer warpage process influence factor information.
Specifically, the production process flow of the target wafer is analyzed based on the wafer warpage loss characteristic information to determine wafer warpage generation factors. Firstly, a wafer production warp database can be obtained through a big data technology, wherein the wafer production warp database is wafer historical production warp data and comprises a wafer warp loss characteristic set and a wafer production process control parameter set. And carrying out simulation training based on the wafer warpage loss feature set and the wafer production process control parameter set to construct a wafer production warpage analyzer, wherein the wafer production warpage analyzer is a neural network model and is used for carrying out production process control parameter analysis of the related influence according to the warpage loss feature.
And analyzing the wafer warpage loss characteristic information to obtain wafer warpage morphological characteristics, namely wafer warpage appearance and wafer warpage degree characteristics, namely wafer warpage degree. And carrying out warpage influence analysis on the wafer warpage morphological characteristics and the wafer warpage characteristics based on the wafer production warpage analyzer, outputting a production process control parameter set which is influenced by the wafer warpage morphological characteristics, and carrying out factor classification and integration on the production process control parameter set to determine wafer warpage process influence factor information, such as process influence factors of cutting, corrosion, annealing and the like of the wafer. And the wafer warping process influence factor information is rapidly analyzed by using the wafer production warping analyzer, so that the factor analysis accuracy and analysis efficiency are improved.
Step S400: performing data searching integration based on the wafer warpage process influence factor information to construct a wafer production strategy database;
step S500: performing similarity matching according to the attribute characteristic information of the target wafer and the wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space;
in one embodiment, the outputting the target wafer production strategy set, step S500 of the present application further includes:
step S510: setting policy feature dimension information according to the attribute feature information;
step S520: classifying and integrating the wafer production strategy database based on the strategy feature dimension information to obtain a wafer production strategy dimension database;
step S530: constructing a wafer production strategy coordinate system, wherein the wafer production strategy coordinate system takes the strategy feature dimension information as a coordinate axis;
step S540: and performing similarity matching on the attribute characteristic information and the wafer production strategy dimension database based on the wafer production strategy coordinate system, and outputting the target wafer production strategy set.
In one embodiment, step S540 of the present application further includes:
step S541: performing regional labeling classification on the wafer production strategy coordinate system based on the wafer production strategy dimension database to obtain a strategy regional labeling classification result;
step S542: inputting the attribute characteristic information into the wafer production strategy coordinate system to obtain a wafer attribute characteristic vector;
step S543: mapping and matching are carried out based on the wafer attribute feature vector and the strategy region labeling classification result, and feature mapping region labels are obtained;
step S544: and determining the target wafer production strategy set according to the feature mapping region label.
Specifically, the wafer production data can be searched and integrated based on the wafer warpage process influence factor information through a data mining technology, and a wafer production strategy database is constructed, wherein the wafer production strategy database is a production control strategy database searched and mined according to the wafer warpage process influence factor information, and comprises production process flows and process control parameters corresponding to each attribute and characteristic type wafer process influence factor. And performing similarity matching with the wafer production strategy database according to the attribute characteristic information of the target wafer, including characteristics such as wafer preparation materials, production sizes and the like. Firstly, according to the attribute feature information, setting strategy feature dimension information, namely, each attribute feature information is used as a strategy feature dimension, wherein the strategy feature dimension information is a production strategy data classification index, is set according to the attribute feature information, comprises feature dimensions such as wafer preparation materials, production sizes and the like, and corresponds to the strategy feature dimension information one by one. And classifying and integrating the wafer production strategy database based on the strategy feature dimension information, namely classifying, integrating and arranging the database according to the strategy feature dimension to obtain an integrated wafer production strategy dimension database.
And constructing a wafer production strategy coordinate system based on the strategy feature dimension information, wherein the wafer production strategy coordinate system takes the strategy feature dimension information as a coordinate axis, and then performing similarity matching on the attribute feature information and the wafer production strategy dimension database based on the wafer production strategy coordinate system. Specifically, the wafer production strategy coordinate system is classified based on the wafer production strategy dimension database in a regional label mode, namely, after data in the database is mapped to the coordinate system, the mapping result is classified into a coordinate system region according to wafer attribute characteristics, production control strategy data mapped by different wafer attribute characteristics are correspondingly different, the coordinate system region is divided into strategy labels corresponding to each wafer attribute characteristic type, and the strategy labels correspond to process influence factors of the wafer production strategy database, including production process flows, process control parameters and the like. The method comprises the steps of establishing a process influence factor type dividing table according to wafer production experience, dividing types of production control strategy data mapped by different wafer attribute characteristics according to the process influence factor type dividing table, determining strategy labels of each wafer attribute characteristic region according to specific types of the process influence factors, and further obtaining corresponding strategy region labeling classification results after the wafer production strategy coordinate system region labeling classification.
And inputting the attribute characteristic information into the wafer production strategy coordinate system, namely carrying out parameter mapping on specific parameters of the attribute characteristic information of the target wafer and all strategy characteristic dimension coordinate axes in the coordinate system, and further obtaining corresponding wafer attribute characteristic vectors through attribute characteristic information mapping values, wherein the wafer attribute characteristic vectors are used for indicating the specific parameters of the attribute characteristic information in the wafer production strategy coordinate system. And mapping and matching are carried out based on the wafer attribute feature vector and the strategy region labeling and classifying result, namely, the mapping region of the wafer attribute feature vector in the wafer production strategy coordinate system is marked, and then strategy label matching corresponding to the strategy region is carried out according to the mapped wafer attribute feature region, and the strategy region labeling and classifying result corresponding to the feature vector matching is obtained, namely, the feature mapping region label. And acquiring strategy data in the wafer production strategy dimension database corresponding to the area label according to the feature mapping area label, and integrating and determining the target wafer production strategy set. And taking the target wafer production strategy set as a wafer warpage optimization space, wherein the wafer warpage optimization space is a wafer warpage production control parameter set capable of being optimally selected. And the wafer warpage optimization space is constructed through similarity matching, so that the optimization parameter optimizing comprehensiveness of the wafer warpage is improved, and the analysis accuracy of the optimization parameter is further improved.
Step S600: constructing a warpage optimization fitness function, performing global optimization in the wafer warpage optimization space based on the warpage optimization fitness function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information;
in one embodiment, the generating wafer warpage optimization parameter information, the applying step S600 further includes:
step S610: randomly selecting first warpage optimization parameter information from the wafer warpage optimization space to serve as an optimal parameter solution;
step S620: calculating the fitness of the optimal parameter solution based on the warping optimization fitness function to obtain first parameter fitness;
step S630: setting a search step length, carrying out parameter search in the wafer warpage optimization space according to the search step length, and calculating second warpage optimization parameter information obtained by searching based on the warpage optimization fitness function to obtain second parameter fitness;
step S640: and if the second parameter fitness is larger than the first parameter fitness, replacing the first warp optimization parameter information with the second warp optimization parameter information to serve as the optimal parameter solution, and outputting the optimal parameter solution until the preset iteration times are reached to obtain the wafer warp optimization parameter information.
Specifically, a warp optimization fitness function is constructed, the warp optimization fitness function can be determined by a plurality of factors such as wafer performance and production cost, the specific determining factors can be set by self, and the warp optimization fitness function is constructed through production and processing experience. And performing global optimization in the wafer warpage optimization space based on the warpage optimization fitness function, and firstly randomly selecting first warpage optimization parameter information from the wafer warpage optimization space to serve as a current optimal parameter solution. And calculating the fitness of the optimal parameter solution based on the warping optimization fitness function to obtain the corresponding first parameter fitness. Setting a parameter searching step length of a wafer warpage optimization space, wherein the smaller the step length is, the larger the searching accuracy is, and carrying out parameter random searching in the wafer warpage optimization space according to the searching step length to obtain second warpage optimization parameter information. And calculating the second warpage optimization parameter information obtained by searching based on the warpage optimization fitness function to obtain the corresponding second parameter fitness.
And comparing the second parameter fitness with the first parameter fitness, and if the second parameter fitness is larger than the first parameter fitness, replacing the first warpage optimization parameter information with the second warpage optimization parameter information to serve as the optimal parameter solution. If the parameter is smaller than the preset parameter, continuing to search and optimize the parameter until the preset iteration times are reached, wherein the preset iteration times can be set according to actual requirements, outputting the optimal parameter solution, and taking the optimal parameter as the wafer warpage optimization parameter information. And carrying out wafer optimized production control according to the wafer warpage optimized parameter information, so that the produced wafer warpage meets the wafer application standard, and improving the accuracy and the comprehensiveness of warpage optimized parameter analysis through parameter global optimization, thereby realizing effective and reliable control and adjustment of the wafer warpage.
In one embodiment, as shown in FIG. 3, a wafer warp optimization system is provided, comprising: the wafer optimizing production control system comprises a warp distribution map acquisition module 11, a warp loss analysis module 12, a process influence factor generation module 13, a strategy database construction module 14, a warp optimizing space acquisition module 15 and a wafer optimizing production control module 16, wherein:
a warp distribution map obtaining module 11, configured to obtain a warp distribution map of a target wafer through a wafer warp stress measurement device;
the warpage loss analysis module 12 is used for acquiring a standard wafer warpage distribution diagram according to a wafer application standard, and carrying out loss analysis on the warpage distribution diagram and the standard wafer warpage distribution diagram based on a digital twin network model to acquire wafer warpage loss characteristic information;
a process influencing factor generating module 13, configured to analyze a production process flow of the target wafer based on the wafer warpage loss feature information, and generate wafer warpage process influencing factor information;
the policy database construction module 14 is configured to perform data search integration based on the wafer warpage process influence factor information, so as to construct a wafer production policy database;
the warpage optimization space obtaining module 15 is configured to perform similarity matching according to the attribute feature information of the target wafer and the wafer production policy database, output a target wafer production policy set, and use the target wafer production policy set as a wafer warpage optimization space;
and the wafer optimization production control module 16 is used for constructing a warpage optimization adaptability function, performing global optimization in the wafer warpage optimization space based on the warpage optimization adaptability function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information.
In one embodiment, the system further comprises:
the digital twin network model building unit is used for building a digital twin network model, and the digital twin network model comprises a target twin network sub-model and a standard twin network sub-model;
the warpage feature set output unit is used for inputting the warpage profile and the standard wafer warpage profile into the target twin network sub-model and the standard twin network sub-model respectively and outputting a target wafer warpage feature set and a standard wafer warpage feature set;
the characteristic loss function acquisition unit is used for acquiring a characteristic loss function, wherein the characteristic loss function is a characteristic comparison value weighted fusion function;
the data loss analysis unit is used for carrying out loss analysis on the target wafer warpage feature set and the standard wafer warpage feature set based on the feature loss function and outputting wafer warpage loss data;
and the feature level classification unit is used for classifying the feature levels of the wafer warpage loss data to obtain the wafer warpage loss feature information.
In one embodiment, the system further comprises:
the model logic layer obtaining unit is used for obtaining an image division logic layer and a feature extraction logic layer according to the target twin network sub-model;
the image grid segmentation unit is used for determining preset division fineness based on the image division logic layer, and carrying out image segmentation on the warpage profile according to the preset division fineness to obtain a grid division warpage profile;
the preset convolution characteristic determining unit is used for determining preset convolution characteristic information of the wafer according to the wafer bending requirement factors;
the traversal convolution calculation unit is used for performing traversal convolution calculation on the grid division warping distribution graph according to the wafer preset convolution characteristic information to obtain the target wafer warping characteristic set conforming to the preset convolution numerical range.
In one embodiment, the system further comprises:
the wafer production warp database comprises a wafer warp loss feature set and a wafer production process control parameter set;
the simulation training unit is used for performing simulation training based on the wafer warpage loss characteristic set and the wafer production process control parameter set to construct a wafer production warpage analyzer;
the loss characteristic analysis unit is used for analyzing the wafer warpage loss characteristic information to obtain wafer warpage morphological characteristics and wafer warpage characteristics;
and the warping influence analysis unit is used for carrying out warping influence analysis on the wafer warping morphological characteristics and the wafer warping degree characteristics based on the wafer production warping analyzer and determining wafer warping process influence factor information.
In one embodiment, the system further comprises:
the policy feature dimension setting unit is used for setting policy feature dimension information according to the attribute feature information;
the database classifying and integrating unit is used for classifying and integrating the wafer production strategy database based on the strategy feature dimension information to obtain a wafer production strategy dimension database;
the strategy coordinate system construction unit is used for constructing a wafer production strategy coordinate system, and the wafer production strategy coordinate system takes the strategy feature dimension information as a coordinate axis;
and the similarity matching unit is used for performing similarity matching on the attribute characteristic information and the wafer production strategy dimension database based on the wafer production strategy coordinate system, and outputting the target wafer production strategy set.
In one embodiment, the system further comprises:
the regional labeling classification unit is used for carrying out regional labeling classification on the wafer production strategy coordinate system based on the wafer production strategy dimension database to obtain a strategy regional labeling classification result;
the attribute feature vector obtaining unit is used for inputting the attribute feature information into the wafer production strategy coordinate system to obtain a wafer attribute feature vector;
the vector mapping matching unit is used for mapping and matching based on the wafer attribute feature vector and the strategy region labeling classification result to obtain a feature mapping region label;
and the production strategy set determining unit is used for determining the target wafer production strategy set according to the feature mapping area tag.
In one embodiment, the system further comprises:
the optimal parameter solution obtaining unit is used for randomly selecting first warpage optimization parameter information from the wafer warpage optimization space to serve as an optimal parameter solution;
a fitness calculating unit, configured to calculate fitness of the optimal parameter solution based on the warpage optimization fitness function, and obtain a first parameter fitness;
the parameter searching unit is used for setting a searching step length, carrying out parameter searching in the wafer warpage optimization space according to the searching step length, and calculating second warpage optimization parameter information obtained by searching based on the warpage optimization adaptability function to obtain second parameter adaptability;
and the wafer warpage optimization parameter information obtaining unit is used for replacing the first warpage optimization parameter information with the second warpage optimization parameter information if the second parameter fitness is larger than the first parameter fitness, and outputting the optimal parameter solution until the preset iteration times are reached, so as to obtain the wafer warpage optimization parameter information.
For a specific embodiment of a wafer warpage optimization system, reference may be made to the above embodiment of a wafer warpage optimization method, which is not described herein. The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.
Claims (8)
1. A method for optimizing wafer warpage, the method comprising:
obtaining a warp distribution map of a target wafer through wafer warp stress measurement equipment;
obtaining a standard wafer warpage distribution map according to a wafer application standard, and carrying out loss analysis on the warpage distribution map and the standard wafer warpage distribution map based on a digital twin network model to obtain wafer warpage loss characteristic information;
analyzing the production process flow of the target wafer based on the wafer warpage loss characteristic information to generate wafer warpage process influence factor information;
performing data searching integration based on the wafer warpage process influence factor information to construct a wafer production strategy database;
performing similarity matching according to the attribute characteristic information of the target wafer and the wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space;
and constructing a warpage optimization fitness function, performing global optimization in the wafer warpage optimization space based on the warpage optimization fitness function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information.
2. The method of claim 1, wherein the obtaining wafer warp loss feature information comprises:
building a digital twin network model, wherein the digital twin network model comprises a target twin network sub-model and a standard twin network sub-model;
inputting the warpage distribution map and the standard wafer warpage distribution map into the target twin network sub-model and the standard twin network sub-model respectively, and outputting a target wafer warpage feature set and a standard wafer warpage feature set;
acquiring a characteristic loss function, wherein the characteristic loss function is a characteristic comparison value weighted fusion function;
performing loss analysis on the target wafer warpage feature set and the standard wafer warpage feature set based on the feature loss function, and outputting wafer warpage loss data;
and classifying the wafer warpage loss data in a characteristic level to obtain the wafer warpage loss characteristic information.
3. The method of claim 2, wherein outputting the set of target wafer warp characteristics comprises:
obtaining an image division logic layer and a feature extraction logic layer according to the target twin network sub-model;
determining a preset division fineness based on the image division logic layer, and performing image segmentation on the warpage profile according to the preset division fineness to obtain a grid division warpage profile;
determining preset convolution characteristic information of the wafer according to the wafer bending requirement factors;
performing traversal convolution calculation on the grid division warping distribution graph according to the wafer preset convolution characteristic information to obtain the target wafer warping characteristic set conforming to a preset convolution numerical range.
4. The method of claim 1, wherein generating wafer warp process influence factor information comprises:
obtaining a wafer production warpage database, wherein the wafer production warpage database comprises a wafer warpage loss feature set and a wafer production process control parameter set;
performing simulation training based on the wafer warpage loss feature set and the wafer production process control parameter set to construct a wafer production warpage analyzer;
analyzing the wafer warpage loss characteristic information to obtain wafer warpage morphological characteristics and wafer warpage characteristics;
and analyzing the warpage influence of the wafer warpage morphological characteristics and the wafer warpage characteristics based on the wafer production warpage analyzer, and determining wafer warpage process influence factor information.
5. The method of claim 1, wherein outputting the set of target wafer production strategies comprises:
setting policy feature dimension information according to the attribute feature information;
classifying and integrating the wafer production strategy database based on the strategy feature dimension information to obtain a wafer production strategy dimension database;
constructing a wafer production strategy coordinate system, wherein the wafer production strategy coordinate system takes the strategy feature dimension information as a coordinate axis;
and performing similarity matching on the attribute characteristic information and the wafer production strategy dimension database based on the wafer production strategy coordinate system, and outputting the target wafer production strategy set.
6. The method of claim 5, wherein the method comprises:
performing regional labeling classification on the wafer production strategy coordinate system based on the wafer production strategy dimension database to obtain a strategy regional labeling classification result;
inputting the attribute characteristic information into the wafer production strategy coordinate system to obtain a wafer attribute characteristic vector;
mapping and matching are carried out based on the wafer attribute feature vector and the strategy region labeling classification result, and feature mapping region labels are obtained;
and determining the target wafer production strategy set according to the feature mapping region label.
7. The method of claim 1, wherein generating wafer warp optimization parameter information comprises:
randomly selecting first warpage optimization parameter information from the wafer warpage optimization space to serve as an optimal parameter solution;
calculating the fitness of the optimal parameter solution based on the warping optimization fitness function to obtain first parameter fitness;
setting a search step length, carrying out parameter search in the wafer warpage optimization space according to the search step length, and calculating second warpage optimization parameter information obtained by searching based on the warpage optimization fitness function to obtain second parameter fitness;
and if the second parameter fitness is larger than the first parameter fitness, replacing the first warp optimization parameter information with the second warp optimization parameter information to serve as the optimal parameter solution, and outputting the optimal parameter solution until the preset iteration times are reached to obtain the wafer warp optimization parameter information.
8. A wafer warp optimization system, the system comprising:
the warp distribution map acquisition module is used for acquiring a warp distribution map of the target wafer through wafer warp stress measurement equipment;
the warpage loss analysis module is used for acquiring a standard wafer warpage distribution diagram according to a wafer application standard, and carrying out loss analysis on the warpage distribution diagram and the standard wafer warpage distribution diagram based on a digital twin network model to acquire wafer warpage loss characteristic information;
the process influence factor generation module is used for analyzing the production process flow of the target wafer based on the wafer warpage loss characteristic information to generate wafer warpage process influence factor information;
the strategy database construction module is used for carrying out data search integration based on the wafer warpage process influence factor information to construct a wafer production strategy database;
the warpage optimization space obtaining module is used for carrying out similarity matching according to the attribute characteristic information of the target wafer and the wafer production strategy database, outputting a target wafer production strategy set, and taking the target wafer production strategy set as a wafer warpage optimization space;
and the wafer optimization production control module is used for constructing a warpage optimization adaptability function, performing global optimization in the wafer warpage optimization space based on the warpage optimization adaptability function, generating wafer warpage optimization parameter information, and performing wafer optimization production control according to the wafer warpage optimization parameter information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311342862.8A CN117094095B (en) | 2023-10-17 | 2023-10-17 | Wafer warpage optimization method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311342862.8A CN117094095B (en) | 2023-10-17 | 2023-10-17 | Wafer warpage optimization method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117094095A true CN117094095A (en) | 2023-11-21 |
| CN117094095B CN117094095B (en) | 2023-12-26 |
Family
ID=88780627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311342862.8A Active CN117094095B (en) | 2023-10-17 | 2023-10-17 | Wafer warpage optimization method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117094095B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117962256B (en) * | 2024-04-01 | 2024-05-31 | 南通世源橡塑科技有限公司 | Injection molding warpage simulation control system based on analysis of big data information |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110427674A (en) * | 2019-07-23 | 2019-11-08 | 杭州晶通科技有限公司 | A kind of warpage emulation mode of fan-out package wafer |
| CN112086350A (en) * | 2020-09-12 | 2020-12-15 | 北京航空航天大学 | Laser grinding process for semiconductor wafer |
| CN115062501A (en) * | 2022-04-27 | 2022-09-16 | 南方科技大学 | Chip packaging design optimization method based on adaptive subproblem selection strategy |
| CN115544956A (en) * | 2022-10-18 | 2022-12-30 | 深圳先进电子材料国际创新研究院 | Evaluation method and evaluation system for thermal mechanical interface stress of chip packaging structure |
-
2023
- 2023-10-17 CN CN202311342862.8A patent/CN117094095B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110427674A (en) * | 2019-07-23 | 2019-11-08 | 杭州晶通科技有限公司 | A kind of warpage emulation mode of fan-out package wafer |
| CN112086350A (en) * | 2020-09-12 | 2020-12-15 | 北京航空航天大学 | Laser grinding process for semiconductor wafer |
| CN115062501A (en) * | 2022-04-27 | 2022-09-16 | 南方科技大学 | Chip packaging design optimization method based on adaptive subproblem selection strategy |
| CN115544956A (en) * | 2022-10-18 | 2022-12-30 | 深圳先进电子材料国际创新研究院 | Evaluation method and evaluation system for thermal mechanical interface stress of chip packaging structure |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117962256B (en) * | 2024-04-01 | 2024-05-31 | 南通世源橡塑科技有限公司 | Injection molding warpage simulation control system based on analysis of big data information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117094095B (en) | 2023-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109977808B (en) | Wafer surface defect mode detection and analysis method | |
| CN109523520B (en) | Chromosome automatic counting method based on deep learning | |
| CN109859207B (en) | Defect detection method of high-density flexible substrate | |
| CN110927171A (en) | Bearing roller chamfer surface defect detection method based on machine vision | |
| CN109284779A (en) | Object detecting method based on the full convolutional network of depth | |
| CN116167668B (en) | BIM-based green energy-saving building construction quality evaluation method and system | |
| WO2021024402A1 (en) | Dimension measurement device, dimension measurement method, and semiconductor manufacturing system | |
| CN113763312A (en) | Detecting defects in semiconductor samples using weak marks | |
| US10345794B2 (en) | Product quality prediction method for mass customization | |
| CN113487533B (en) | Part assembly quality digital detection system and method based on machine learning | |
| CN115100199A (en) | Method for detecting wafer low-texture defects | |
| CN112365497A (en) | High-speed target detection method and system based on Trident Net and Cascade-RCNN structures | |
| CN114910480A (en) | Wafer surface defect detection method based on machine vision | |
| CN110716500A (en) | Method and system for determining segmented modeling points of temperature sensitive interval | |
| CN116128839A (en) | Wafer defect identification method, device, electronic equipment and storage medium | |
| CN115147363A (en) | Image defect detection and classification method and system based on deep learning algorithm | |
| CN117094095B (en) | Wafer warpage optimization method and system | |
| CN116433661B (en) | Method, device, equipment and medium for detecting semiconductor wafer by multitasking | |
| CN116205918B (en) | Multi-mode fusion semiconductor detection method, device and medium based on graph convolution | |
| CN115082444B (en) | Copper pipe weld defect detection method and system based on image processing | |
| CN110675363A (en) | Automatic calculation method of DNA index for cervical cells | |
| CN110533657A (en) | A kind of liquid crystal display appearance detecting method | |
| CN115330705A (en) | Skin paint surface defect detection method based on adaptive weighting template NCC | |
| CN114911085A (en) | Method for analyzing defects | |
| Yu et al. | A Machine vision method for non-contact Tool Wear Inspection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |