CN116230586A - Commonality analysis method and terminal of wafer manufacturing machine unit - Google Patents
Commonality analysis method and terminal of wafer manufacturing machine unit Download PDFInfo
- Publication number
- CN116230586A CN116230586A CN202211591455.6A CN202211591455A CN116230586A CN 116230586 A CN116230586 A CN 116230586A CN 202211591455 A CN202211591455 A CN 202211591455A CN 116230586 A CN116230586 A CN 116230586A
- Authority
- CN
- China
- Prior art keywords
- wafer
- abnormal
- correlation
- value
- analyzed
- 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
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 93
- 238000004458 analytical method Methods 0.000 title claims abstract description 65
- 235000012431 wafers Nutrition 0.000 claims abstract description 395
- 238000000034 method Methods 0.000 claims abstract description 166
- 230000002159 abnormal effect Effects 0.000 claims abstract description 163
- 238000012360 testing method Methods 0.000 claims abstract description 58
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000010606 normalization Methods 0.000 claims abstract description 12
- 239000013598 vector Substances 0.000 claims description 96
- 238000004590 computer program Methods 0.000 claims description 10
- 230000002688 persistence Effects 0.000 claims description 9
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000010219 correlation analysis Methods 0.000 abstract description 7
- 238000004445 quantitative analysis Methods 0.000 abstract description 5
- 230000008569 process Effects 0.000 description 22
- 238000000540 analysis of variance Methods 0.000 description 18
- 230000005856 abnormality Effects 0.000 description 16
- 238000004422 calculation algorithm Methods 0.000 description 15
- 238000004364 calculation method Methods 0.000 description 8
- 238000003860 storage Methods 0.000 description 8
- 230000007547 defect Effects 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L21/00—Processes or apparatus adapted for the manufacture or treatment of semiconductor or solid state devices or of parts thereof
- H01L21/67—Apparatus specially adapted for handling semiconductor or electric solid state devices during manufacture or treatment thereof; Apparatus specially adapted for handling wafers during manufacture or treatment of semiconductor or electric solid state devices or components ; Apparatus not specifically provided for elsewhere
- H01L21/67005—Apparatus not specifically provided for elsewhere
- H01L21/67242—Apparatus for monitoring, sorting or marking
- H01L21/67276—Production flow monitoring, e.g. for increasing throughput
-
- 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)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- General Physics & Mathematics (AREA)
- Manufacturing & Machinery (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Power Engineering (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
- General Factory Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A method and a terminal for analyzing the commonality of a wafer manufacturing machine unit, wherein the method comprises the following steps: calculating a correlation degree based on the wafer data, wherein the correlation degree is used for representing the correlation degree between the dimension to be analyzed and the target test item; determining a basic value at least according to an abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence; carrying out normalization processing on the correlation degree based on the base value to obtain a correlation coefficient; and obtaining an analysis result of the dimension to be analyzed based on the value of the correlation coefficient. According to the technical scheme, the accuracy of the commonality analysis of the machine table/machine table chamber can be improved, the quantitative analysis of the commonality correlation can be further realized, and the commonality correlation analysis can be accurately carried out when a specific process step is executed by a single wafer manufacturing machine table unit and the process step is problematic.
Description
Technical Field
The invention relates to the technical field of chip manufacturing and testing, in particular to a commonality analysis method and a terminal of a wafer manufacturing machine unit.
Background
The wafer fabrication process flow includes thousands of process steps, any of which may be problematic and may result in poor quality, high defects, or any abnormal wafer test parameters. When the engineer performs failure analysis, it finds out which process machine of which process step is the root cause and even which process parameter of the machine is problematic according to various correlation analysis methods.
At this stage, tool Commonality (Tool Commonality) and Tool chamber Commonality (Chamber Commonality) are the most common analysis methods. Through analysis of variance (Analysis of Variance, ANOVA) algorithms, engineers want to find out whether there is a lot of bad, high defect, or test parameter anomaly, or whether there is a difference in good or bad tools. For example, in analyzing a low-yield case where the test parameter a fails, an ANOVA analysis finds that there is a significant difference between the yield loss of the test parameter a for the lot or wafer of the tool 01 that has undergone a certain process step and the other tools 02, 03 that have undergone that process step. At this time, the machine 01 capable of locking this process step is the root cause of low yield loss of the test parameter a.
However, when the case analysis is actually performed, the situation that the machine commonality analysis cannot find the machine with the difference of quality or the root cause machine causing the case cannot be found is often encountered. Moreover, ANOVA analysis itself has algorithm defects, and may mislead engineers to fail analysis results that are inconsistent with the actual situation.
Disclosure of Invention
The invention solves the technical problem of how to improve the accuracy of the commonality analysis of the machine/machine cavity.
In order to solve the above technical problems, an embodiment of the present invention provides a method for analyzing commonality of a wafer manufacturing tool unit, including: calculating correlation based on wafer data, wherein the correlation is used for representing the correlation degree of the dimension to be analyzed and the target test item, and the wafer data is data of related wafers obtained based on the process steps of the production flow; determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence; carrying out normalization processing on the correlation degree based on the base value to obtain a correlation coefficient; based on the value of the correlation coefficient, obtaining an analysis result of the dimension to be analyzed; wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
Optionally, the calculating the correlation based on the wafer data includes: statistically obtaining abnormal wafer continuous values associated with the dimension to be analyzed from the wafer data; determining the length of a base vector at least according to the abnormal wafer continuous value, wherein the base vector comprises elements for representing that the state of the wafer is abnormal; extracting an abnormal wafer continuous vector associated with the dimension to be analyzed from the wafer data, wherein the abnormal wafer continuous vector comprises elements at least used for representing the states of all wafers from the first abnormal wafer to the last abnormal wafer which are sequenced according to the production time in the wafer data, and the length of the abnormal wafer continuous vector is consistent with the length of the base vector; and calculating the similarity between the abnormal wafer continuous vector and the base vector to obtain the correlation.
Optionally, the determining the length of the base vector at least according to the abnormal wafer duration value includes: counting the total number of abnormal wafers associated with the process steps in the wafer data; and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the length of the base vector.
Optionally, the extracting the abnormal wafer duration vector associated with the dimension to be analyzed from the wafer data includes: if the number of the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence is smaller than the length of the base vector, filling preset elements in the abnormal wafer continuous vector so that the length of the abnormal wafer continuous vector is consistent with the length of the base vector.
Optionally, the calculating the correlation based on the wafer data includes: and counting the total number of abnormal wafers associated with the dimension to be analyzed in the wafer data to obtain the correlation degree.
Optionally, the determining the base value at least according to the abnormal wafer duration value associated with the dimension to be analyzed includes: counting the total number of abnormal wafers associated with the process steps in the wafer data; and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the base value.
Optionally, the wafer data includes a plurality of records, wherein each record includes an identification of a wafer and a test value of the wafer at the target test item, and the method further includes: for each record, marking the state of the wafer according to the preset threshold value of the target test item and the test value, wherein the state comprises abnormal and normal states.
Optionally, for the correlation coefficient of the same process step under a plurality of dimensions to be analyzed, taking the correlation coefficient with the largest value as the correlation coefficient of the process step.
Optionally, the wafer manufacturing tool unit includes a wafer manufacturing tool or a chamber of the wafer manufacturing tool.
Optionally, the obtaining the analysis result of the dimension to be analyzed based on the value of the correlation coefficient includes: sequencing the dimension to be analyzed of all the process steps according to the numerical value of the correlation coefficient to obtain an analysis result; or sequencing the dimensions to be analyzed of all the process steps according to the numerical value of the correlation coefficient and a preset coefficient to obtain an analysis result, wherein the preset coefficient is related to the dimensions to be analyzed.
In order to solve the above-mentioned technical problems, an embodiment of the present invention provides a commonality analysis device of a wafer manufacturing machine unit, including: the first processing module is used for calculating correlation degree based on wafer data, wherein the correlation degree is used for representing the correlation degree between the dimension to be analyzed and the target test item, and the wafer data are data of related wafers obtained based on the process steps of the production flow; the determining module is used for determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data, which are ordered according to the production time; the second processing module is used for carrying out normalization processing on the correlation degree based on the basic value to obtain a correlation coefficient; the third processing module is used for obtaining an analysis result of the dimension to be analyzed based on the value of the correlation coefficient; wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
To solve the above technical problem, embodiments of the present invention further provide a computer readable storage medium, where the computer readable storage medium is a non-volatile storage medium or a non-transitory storage medium, and a computer program is stored thereon, and when the computer program is executed by a processor, the steps of the above method are performed.
In order to solve the above technical problem, an embodiment of the present invention further provides a terminal, which includes a memory and a processor, where the memory stores a computer program that can be run on the processor, and the processor executes the steps of the above method when running the computer program.
Compared with the prior art, the technical scheme of the embodiment of the invention has the following beneficial effects:
the embodiment of the invention provides a commonality analysis method of a wafer manufacturing machine station unit, which comprises the following steps: calculating correlation based on wafer data, wherein the correlation is used for representing the correlation degree of the dimension to be analyzed and the target test item, and the wafer data is data of related wafers obtained based on the process steps of the production flow; determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence; carrying out normalization processing on the correlation degree based on the base value to obtain a correlation coefficient; based on the value of the correlation coefficient, obtaining an analysis result of the dimension to be analyzed; wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
Compared with the existing analysis scheme for the commonality correlation based on the ANOVA algorithm, the method and the device for processing the commonality correlation based on the ANOVA algorithm process based on the correlation, and the factor of continuity of wafer abnormality is combined to process the correlation, so that the finally obtained correlation coefficient can more accurately represent the commonality correlation of a machine/machine chamber/process step and a target test item. Specifically, the correlation coefficient is obtained after normalization by using a base value, and the base value is determined according to the abnormal wafer continuous value, so that the correlation coefficient can reflect the difference of the correlation of different dimensions to be analyzed on the continuity of wafer abnormality, such as the influence of intermittent abnormality and continuous abnormality in a period of time on the correlation calculation result. Thus, the accuracy of the commonality correlation analysis can be greatly improved.
Furthermore, the present embodiment can realize quantitative analysis of the commonality correlation, and precisely quantify the correlation between the process steps or wafer manufacturing tool units and the target test parameters.
Further, the embodiment can solve the problem of a single machine or a certain process step which cannot be solved by the prior art. Specifically, since the analysis of the commonality correlation in this embodiment is not dependent on the differences between the multiple tools of the same process step, but rather quantifies the correlation coefficient of each tool/chamber of each process step, the analysis of the commonality correlation can still be accurately performed when a particular process step is performed by a single wafer fabrication tool unit. Further, since the analysis dimension of the present embodiment is no longer limited to the differences between the machines, the analysis of the commonality correlation can still be accurately performed when problems occur in the process steps themselves.
Drawings
FIG. 1 is a flow chart illustrating a method for analyzing commonality of wafer fabrication tool units according to an embodiment of the present invention;
FIG. 2 is a flow chart of one embodiment of step S101 of FIG. 1;
FIG. 3 is a diagram showing the correlation of a specific machine for a specific process step in a typical application scenario according to an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating a structure of a commonality analysis apparatus of a wafer manufacturing tool unit according to an embodiment of the present invention.
Detailed Description
As described in the background art, the existing method for analyzing the commonality of the machine/machine chamber has the problem of inaccurate analysis results.
Specifically, the existing commonality analysis method is generally realized based on an ANOVA algorithm. The ANOVA algorithm is essentially to count the variance of the wafer of each machine after a certain process step on a certain test parameter, and compare the variance of each machine.
For example, assume that there are two process steps, respectively denoted as step 01 and step 02, where step 01 is failed in testing parameter a for 100 wafers in an intermittent manner for 300 wafers passing through the machine 01, and step 01 is failed in testing parameter a for 500 wafers passing through the machine 02; step 02, testing parameters A of 100 continuous wafers in 700 wafers passing through a 03 machine table are failed, and step 02, no wafer testing parameters A in 100 wafers passing through a 04 machine table are failed. Because the difference between the failure rate 100/300 of the machine 01 and the failure rate 0 of the machine 02 in the step 01 is higher than the difference between the failure rate 100/700 of the machine 03 and the failure rate 0 of the machine 02 in the step 02, according to the existing ANOVA algorithm, the machine variance of the step 01 is calculated to be larger, so that the correlation between the step 01 and the test parameter A is considered to be higher. But in practice it should be that the commonality of the problem with step 02 is higher.
The inventor of the application found through analysis that one of the reasons for the problems is that the conventional analysis of the commonality correlation based on the ANOVA algorithm does not consider the factor of the continuity of wafer abnormality, so that the analysis result is inaccurate and one-sided exists. The existing commonality correlation analysis is to count the deviation degree of the test parameters from the mean value after the wafer passes through the machine on the whole, and for a certain machine, the higher the test parameter reject ratio of the wafer passing through the machine in a case is, the higher the commonality correlation of the machine and the test parameters is considered to be problematic. In fact, however, the intermittent occurrence of a problem with a wafer passing through a certain station and the occurrence of a problem with all wafers passing through the station over a period of time are of very different importance to engineers, the latter being obviously more serious to be handled as early as possible. However, as exemplified above, the existing ANOVA algorithm cannot determine the machine with small variance but continuously with bad chips as high correlation, and therefore, the machine cannot be timely prompted to an engineer, and in severe cases, the engineer may be misled to obtain failure analysis results inconsistent with actual situations.
In addition, the existing machine/machine cavity commonality analysis method also often encounters the situation that the machine with the difference of quality cannot be found in the machine commonality analysis or the root cause machine of the case cannot be found. The reasons that may lead to these situations are:
1. For throughput, in most cases, the same process step is performed not by one machine, but by a plurality of machines, and if there are machine differences in the process steps, the faulty machine is likely to be grasped out by the ANOVA algorithm. However, in actual production, many process steps are performed by a single machine, i.e., only one machine can perform the process steps. If the problem occurs in a single machine, the difference of good and bad machines does not exist, and the existing ANOVA algorithm is naturally ineffective.
2. Some cases may be due to problems with certain process steps during certain time periods. That is, although the process step has a plurality of machines that can produce the process step, it is not a problem with a certain machine, but any machine that has passed the process step for a certain period of time. At this time, the ANOVA algorithm finds that the difference between all the machines is very small, that is, there is no difference between good and bad machines, and all the machines pass through both good and bad lots, so that the ANOVA algorithm is naturally ineffective.
In order to solve the above technical problems, an embodiment of the present invention provides a method for analyzing commonality of a wafer manufacturing tool unit, including: calculating correlation based on wafer data, wherein the correlation is used for representing the correlation degree of the dimension to be analyzed and the target test item, and the wafer data is data of related wafers obtained based on the process steps of the production flow; determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence; carrying out normalization processing on the correlation degree based on the base value to obtain a correlation coefficient; based on the value of the correlation coefficient, obtaining an analysis result of the dimension to be analyzed; wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
By adopting the embodiment, the correlation degree is processed by combining the factor of the continuity of the wafer abnormality on the basis of calculating the correlation degree, so that the finally obtained correlation coefficient can more accurately represent the commonality correlation of the machine table/machine table chamber/process step and the target test item. Specifically, the correlation coefficient is obtained after normalization by using a base value, and the base value is determined according to the abnormal wafer continuous value, so that the correlation coefficient can reflect the difference of the correlation of different dimensions to be analyzed on the continuity of wafer abnormality, such as the influence of intermittent abnormality and continuous abnormality in a period of time on the correlation calculation result. Thus, the accuracy of the commonality correlation analysis can be greatly improved. Furthermore, the present embodiment can realize quantitative analysis of the commonality correlation, and precisely quantify the correlation between the process steps or wafer manufacturing tool units and the target test parameters. Further, since the analysis of the commonality correlation in the present embodiment is not dependent on the differences between the different tools in the same process step, but quantifies the correlation coefficient of each tool/chamber in each process step, the analysis of the commonality correlation can be accurately performed when a specific process step is performed by a single wafer manufacturing tool unit. Further, since the analysis dimension of the present embodiment is no longer limited to the differences between the machines, the analysis of the commonality correlation can still be accurately performed when problems occur in the process steps themselves.
In order to make the above objects, features and advantages of the present invention more comprehensible, embodiments accompanied with figures are described in detail below.
FIG. 1 is a flow chart illustrating a method for analyzing commonality of wafer fabrication tool units according to an embodiment of the present invention.
The embodiment can be applied to a chip failure analysis scene. The commonality analysis method may be performed by a device performing or assisting in the testing, such as may be performed by an automated testing device (Automatic Test Equipment, ATE for short) or a computing device coupled locally or remotely to the ATE.
Specifically, the embodiment may be used to determine the commonality correlation between the wafer fabrication tool unit and the target test item, and quantify the correlation analysis result by the correlation coefficient. The correlation coefficient with higher accuracy is obtained by executing the embodiment, which is helpful for engineers to accurately analyze the root cause of wafer abnormality, and further is helpful for improving the reliability of the chip.
The "wafer fabrication tool unit" in this embodiment may include a wafer fabrication tool (which may be referred to as a tool or a process tool). The wafer fabrication tool may include one or more chambers (also referred to as chambers) each for processing wafers in a particular process step, and thus the wafer fabrication tool unit in this embodiment may also include chambers.
The "target test item" in this embodiment may include one or more test parameters/items of the wafer, such as pin (bin) 3, bin1 and bin7, bin5, etc. In some embodiments, the target test item may be associated with a case (case), where wafer data related to the same case is used to analyze the root cause of failure of a wafer manufactured in a period of time under the same target test item.
A single case may relate to wafer data for thousands of wafers, i.e., the wafer data may include a plurality of records, where each record may include an identification of a wafer and a test value under a target test item. For example, the identification of the WAFER may include a LOT number (lot_id) and a WAFER number (wafer_no), and a WAFER may be uniquely identified by the cooperation of the LOT number and the WAFER number.
More specifically, referring to fig. 1, the method for analyzing the commonality of the wafer manufacturing tool unit according to the present embodiment may include the following steps:
step S101, calculating correlation based on wafer data, wherein the correlation is used for representing the correlation degree of the dimension to be analyzed and the target test item, and the wafer data is data of related wafers obtained based on the process steps of the production flow;
Step S102, determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence;
step S103, carrying out normalization processing on the correlation degree based on the basic value to obtain a correlation coefficient;
step S104, based on the value of the correlation coefficient, an analysis result of the dimension to be analyzed is obtained.
Further, the dimension to be analyzed may be at least selected from: and a wafer manufacturing machine unit for executing the process steps. That is, during the execution of steps S101 to S103, each process step may be analyzed by wafer fabrication tool units, or may not be analyzed by wafer fabrication tool units, but all wafer fabrication tool units of the same process step may be put together for analysis.
In some embodiments, the dimension to be analyzed may be a wafer manufacturing tool unit performing the process step, and the wafer data in step S101 may include a record of the wafer after passing through the specific wafer manufacturing tool unit. Thus, analyzing whether a wafer has passed through a wafer fabrication tool unit within a certain period of time is problematic.
In some embodiments, the dimension to be analyzed may be a process step, and the wafer data in step S101 may include a record of the wafers processed by all the wafer manufacturing tools performing the process step. Thus, there is no consideration of the respective correlations of the different wafer fabrication tools of the same process step in order to analyze whether a wafer having undergone a certain process step within a certain time period is problematic.
In one implementation, for each process step, steps S101 to 103 may be performed from the process step and the wafer fabrication tool unit performing the process step, respectively, to achieve quantitative analysis of the correlation of the tool and process step commonalities.
Specifically, for a plurality of dimensions to be analyzed, steps S101 to S103 may be performed independently for each to obtain corresponding correlation coefficients, respectively. For example, for a particular process step, the apparatus performing the present embodiment may calculate a process tool correlation and a process step correlation, respectively.
Further, for the correlation coefficients of the same process step in a plurality of dimensions to be analyzed, in step S104, the correlation coefficient in which the numerical value is the largest may be taken as the correlation coefficient of the process step. For example, for a certain process step, assuming that the correlation coefficient with the process machine is calculated to be 0.7 and the correlation coefficient with the process step is calculated to be 0.9, the correlation coefficient of the process step is determined to be the common correlation of the process step and participates in the sorting action in step S104, and is presented as the final analysis result.
In one implementation, the whole production process may include N process steps, and the correlation coefficient of the N process steps may be calculated by performing steps S101 to S103. Further, steps S101 to S103 may be repeatedly performed N times to traverse all process steps of the entire production flow. N may be, for example, 800, n=1, 2,3, … …,800.
In some embodiments, a multi-threaded process may be employed to accelerate traversal through the N process steps. For example, a plurality of threads are run in parallel, wherein each thread runs the scheme of step S101 to step S103, and different threads are used to calculate correlation coefficients of dimensions to be analyzed of different process steps. Finally, the calculation results of all threads are merged together at step S104.
Therefore, the quantitative operation of the commonality correlation in the embodiment supports multithreading to rapidly cover the whole manufacturing production flow, and is beneficial to improving the analysis efficiency.
In one implementation, in step S101, the test values recorded in each piece of wafer data may be first explicit, and then the correlation may be calculated based on the explicit result.
Dominance may specifically refer to making a wafer involved in a case dominant with a bad-not-good identifier. The explicit result of binary classification instead of the test value is used as the basis of the subsequent correlation calculation, so that the calculation complexity in the correlation calculation is reduced.
The specific implementation of the explicit process may specifically include: for each record, marking the state of the wafer according to the preset threshold value of the target test item and the test value, wherein the state comprises abnormal and normal states. The wafer in abnormal state is a Bad wafer (indicated by 1 or Bad), and the wafer in normal state is a Good wafer (indicated by 0 or Good).
During failure analysis, the index of each wafer related to the case is usually represented by a numerical value, such as an average value of a certain test parameter of each wafer, a Yield Loss (Bin Yield Loss) of a certain test item, and the number of defects, which are the test values of the target test item. In this embodiment, for each record in the wafer data, a list of status information is added to the existing test values to identify the status of the wafer for that record.
In some embodiments, the preset threshold for measuring the quality of the wafer may be set by the user according to the need, for example, the magnitude of the preset threshold may be adjusted according to the tolerance for the bad wafer.
Table 1 shows an exemplary list of the good and bad wafers of the case part, which is the dominant wafer data in step S101:
TABLE 1
The label (label) represents the state of the wafer, and the target test item is yield loss. One row in table 1 is a record corresponding to information about a wafer.
In one implementation, for each record in the wafer data, the record may further include the process steps the wafer was subjected to and the wafer fabrication tool unit.
In one implementation, the relevance may be used to characterize the relevance of the dimension to be analyzed and the case.
Specifically, referring to fig. 2, step S101 may include the steps of:
step S1011, statistically obtaining abnormal wafer continuous values associated with the dimension to be analyzed from the wafer data;
step S1012, determining a length of a base vector at least according to the abnormal wafer duration value, where the base vector includes elements for characterizing that the state of the wafer is abnormal;
step S1013, extracting an abnormal wafer duration vector associated with the dimension to be analyzed from the wafer data, where the abnormal wafer duration vector includes elements at least used to characterize states of each wafer from a first abnormal wafer to a last abnormal wafer in the wafer data, where the first abnormal wafer to the last abnormal wafer are ordered according to production time, and the length of the abnormal wafer duration vector is consistent with the length of the base vector;
Step S1014, calculating the similarity between the abnormal wafer duration vector and the base vector, to obtain the correlation.
In order to quantify the change of the wafer data, the continuity characteristic of the variables is embodied by using vectors, so that the implementation converts the wafer data into vector representation and utilizes the vector similarity to represent the dimension to be analyzed and the relevance of the case.
In some embodiments, it may be assumed that the wafers continuously passing through the dimension to be analyzed are all bad to obtain the basis vector. That is, all elements in the base vector are 1.
In some embodiments, the length of the basis vector may be consistent with the basis value determined in step S102. That is, the number of elements included in the base vector affects the mapping radix at the time of the subsequent normalization processing.
In one implementation, all good and bad wafers in a case may be grouped according to the dimension to be analyzed of the current process step, and the wafers may be sorted according to their respective production times for each dimension to be analyzed.
Taking the dimension to be analyzed as a table as an example, table 2 exemplarily shows a list of part of wafers of the table 01 (numbered HFXXX 01) of the step1 process step, ordered by production time:
TABLE 2
Where Step1 represents the 1 st process Step and the production Time is represented based on the tracking Time (track_in_time).
Table 3 illustrates a list of part of wafers of stage 02 (numbered HFXXX 02) for step1 process step, ordered by production time:
TABLE 3 Table 3
Batch number | Wafer number | Label (Label) | Process steps | Machine table | Production time |
Lot05 | 7 | 1 | Step1 | HFXXX02 | 2022/6/11 6:02:07 |
|
4 | 0 | Step1 | HFXXX02 | 2022/6/11 6:02:51 |
Lot04 | 5 | 0 | Step1 | HFXXX02 | 2022/6/11 9:26:32 |
Lot04 | 6 | 0 | Step1 | HFXXX02 | 2022/6/11 9:41:18 |
Lot04 | 7 | 0 | Step1 | HFXXX02 | 2022/6/11 10:10:39 |
Lot04 | 8 | 0 | Step1 | HFXXX02 | 2022/6/13 14:02:06 |
Lot04 | 9 | 0 | Step1 | HFXXX02 | 2022/6/13 16:20:41 |
Lot04 | 10 | 0 | Step1 | HFXXX02 | 2022/7/21 17:46:19 |
Lot06 | 8 | 1 | Step1 | HFXXX02 | 2022/7/21 21:24:47 |
Lot06 | 9 | 1 | Step1 | HFXXX02 | 2022/7/21 22:27:22 |
Lot06 | 10 | 1 | Step1 | HFXXX02 | 2022/7/21 23:11:40 |
Where Step1 represents the 1 st process Step and the production Time is represented based on the tracking Time (track_in_time). It should be noted that tables 1, 2 and 3 are all partial records of all wafer data related to a case, and the records shown in tables 2 and 3 are not completely disclosed in table 1.
Further, in each dimension to be analyzed, taking each set of process tools as an example, the first bad wafer (i.e., the first abnormal wafer) and the last bad wafer (i.e., the last abnormal wafer) in the set of process tools are found out according to the production time sequence. Wherein one machine for producing one process step is denoted as a set of process machines. For example, the first bad Wafer in table 2 is Lot05, wafer 1, and the last bad Wafer is Lot05, wafer 6. For another example, the first bad Wafer in table 3 is Lot05, wafer 7, and the last bad Wafer is Lot06, wafer 10.
Therefore, the abnormal duration condition related to the dimension to be analyzed in all the wafers related to the case can be obtained by determining the first abnormal wafer and the last abnormal wafer which are grouped according to the dimension to be analyzed and are ordered according to the production time in the wafer data. The anomaly persistence is characterized based on anomaly wafer persistence values, which may be measured by the number of all state wafers located between the first anomaly wafer (inclusive) and the last anomaly wafer in the production time order.
In one implementation, step S1012 may directly determine the abnormal wafer duration value as the length of the basis vector when determining the length of the basis vector.
In one implementation, in addition to the abnormal wafer duration values, the total number of abnormal wafers associated with the process steps may be combined to comprehensively determine a basis vector having a more suitable length.
Specifically, the total number of abnormal wafers may be the total number of wafers with the dominance of the target test item of 1 after the process step.
Further, step S1012 may include the steps of: counting the total number of abnormal wafers associated with the process steps in the wafer data; and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the length of the base vector.
That is, if the abnormal wafer duration value is smaller than the total number of abnormal wafers, determining the total number of abnormal wafers as the length of the base vector; otherwise, the abnormal wafer duration value is determined as the length of the basis vector.
Therefore, the persistence and the total bad film condition can be combined, and the length of the base vector can be reasonably determined. Further, by making the length of the base vector be the larger value of the abnormal wafer duration value and the total number of abnormal wafers, it is possible to ensure that the length of the base vector is not too short, that is, that the base value as the normalization base in step S102 is not too small.
In one implementation, the determination of the base value in step S102 may refer to the determination of the length of the base vector in the above embodiment. The method specifically comprises the following steps: counting the total number of abnormal wafers associated with the process steps in the wafer data; and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the base value.
Specifically, if the abnormal wafer duration value is smaller than the total number of abnormal wafers, determining the total number of abnormal wafers as a base value; otherwise, the abnormal wafer duration value is determined as a base value.
For example, assume that the dimension to be analyzed is the chamber 02 of the tool 01 performing the 5 th process step, and after statistics, 8 abnormal wafers are present in total (i.e., the total number of abnormal wafers is 8) in all the tools of the 5 th process step, wherein 3 abnormal wafers are present in the chamber 02 of the tool 01 (i.e., the duration value of the abnormal wafers is 3). In this example, the base value takes the larger of the total number of abnormal wafers and the duration value of the abnormal wafers, i.e., 8.
In one implementation, in step S1013, the states of the wafers from the first abnormal wafer to the last abnormal wafer, which are associated with the dimension to be analyzed and are ordered according to the production time, in the wafer data may be used as the elements of the vector, and the abnormal wafer duration vector may be extracted.
Specifically, the arrangement sequence of the elements in the abnormal wafer duration vector is consistent with the production time sequence of the wafers respectively represented by the elements in the wafer data.
Further, the length of the abnormal wafer duration vector may be consistent with the length of the base vector, so as to calculate the vector similarity in step S1013.
The possibility that the length of the abnormal wafer duration vector extracted from the wafer data is smaller than the length of the base vector is considered, and the possibility is considered to occur when the abnormal wafer duration value is smaller than the total number of abnormal wafers. Thus step S1013 may include the steps of: if the number of the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence is smaller than the length of the base vector, filling preset elements in the abnormal wafer continuous vector so that the length of the abnormal wafer continuous vector is consistent with the length of the base vector.
For example, 0 may be added at the end of the extracted abnormal wafer duration vector so that the length of the abnormal wafer duration vector is equal to the total number of abnormal wafers, that is, the length of the base vector. Note that the 0-complement operation in this example is because the state of a good wafer is defined as 0 and the state of an abnormal wafer is defined as 1. In practical applications, the opposite definition may be adopted, i.e. the good wafer is 1 and the bad wafer is 0, and at this time, 1 may be added at the end of the abnormal wafer duration vector to add length.
In one implementation, in step S104, the dimensions to be analyzed of all the process steps may be sorted according to the magnitude of the correlation coefficient, so as to obtain an analysis result. In particular, a higher score for the correlation coefficient means a higher commonality correlation for the case and the dimension to be analyzed for the corresponding process step, which facilitates the engineer's rapid localization analysis by ordering and displaying the dimension to be analyzed of the highest correlation in front.
Further, for multiple dimensions to be analyzed of the same process step, the dimension to be analyzed in which the value of the correlation coefficient is the largest may be highlighted. This also helps engineers to quickly locate and follow-up analysis.
In a variation, in step S104, the dimensions to be analyzed of all the process steps may be sorted according to the magnitude of the value of the correlation coefficient and a preset coefficient, so as to obtain an analysis result, where the preset coefficient is associated with the dimensions to be analyzed.
Specifically, on the basis of calculating the correlation coefficient of each dimension to be analyzed in each process step, the corresponding correlation coefficient is weighted based on the preset coefficient, so that the finally participated numerical value accords with the actual situation of the case.
Further, the preset coefficient may be a value of 0 to 1 or a value of 1 or more.
In some embodiments, the preset coefficients may relate to factors such as importance of the dimension to be analyzed, whether it is a preparation key node, whether it is a new process/new machine, a common relevance ranking obtained by historical analysis, and the like. The preset coefficients of the dimensions to be analyzed are reasonably designed, so that the commonality correlation of the dimensions to be analyzed can be more reasonably determined.
In some embodiments, the dimensions to be analyzed of the same process step may be associated with the same preset coefficient. Further, the preset coefficients associated with the dimensions to be analyzed of at least one process step are different from the preset coefficients associated with the dimensions to be analyzed of other process steps.
In some embodiments, for multiple dimensions to be analyzed of the same process step, the preset coefficients associated with at least one dimension to be analyzed may be different from the preset coefficients associated with other dimensions to be analyzed.
In a typical application scenario, a base vector may be defined first, and still taking the wafer data shown in tables 2 and 3 as an example, there are 10 bad wafers in the two tables, i.e. the total number of abnormal wafers is 10. At this time, the base vector is defined as a 1×10 vector [1,1,1,1,1,1,1,1,1,1 ] in which all elements are 1] 1×10 。
The number of the first to last bad wafers in table 2 is 6, which is smaller than the dimension 10 of the base vector, so the base vector is used as the base vector corresponding to the Step1 HFXXX01 machine. In Table 3, the first bad wafer to the last bad wafer have 11 wafers, and the number of the bad wafers is greater than the dimension 10 of the base vector, and the corresponding base vector of Step1 HFXXX02 is [1,1,1,1,1,1,1,1,1,1,1 ] ] 1×11 。
And then, converting the good/bad labels of all the wafers from the first bad wafer (including) to the last bad wafer (including) of each sequenced process machine into a one-dimensional vector to obtain an abnormal wafer continuous vector. If the dimension of the abnormal wafer continuous vector is smaller than that of the base vector, the corresponding number of 0 s is complemented at the last of the abnormal wafer continuous vector. The data in table 2 is converted to an abnormal wafer duration vector followed by [1,1,1,1,1,1,0,0,0,0], and the data in table 3 is converted to an abnormal wafer duration vector followed by [1,0,0,0,0,0,0,0,1,1,1].
And then, comparing the abnormal wafer continuous vector converted by each process machine with the corresponding base vector, and taking the calculated vector similarity as a correlation degree. For example, the similarity between the abnormal wafer persistence vector and the basis vector after the Step1 HFXXX01 process is 6, and the similarity between the abnormal wafer persistence vector and the basis vector after the Step1 HFXXX02 process is 4.
And finally, normalizing the calculated correlation degree, so that the normalized correlation coefficient can reflect the proportion of the abnormal wafers associated with the dimension to be analyzed to the total abnormal wafers of the case on the basis of considering the continuity. For example, step1 HFXXX01 machines have a similarity of 6/10= 0.6,Step1 HFXXX02 machines and 4/11= 0.3636 machines.
In one implementation, after traversing all process steps of the overall production flow, step S104 may be performed to rank all calculated correlation coefficients from high to low in value. The higher the value of the correlation coefficient is, the higher the correlation between the dimension to be analyzed and the case is.
Taking the dimension to be analyzed as a machine as an example, table 4 exemplarily shows the Correlation coefficient (Correlation) of the front 10 of the numerical value row obtained by final calculation and the corresponding machine and process steps:
TABLE 4 Table 4
It can be concluded from table 4 that the HFXXR02 machine with the highest correlation coefficient of Step31 corresponds to a correlation coefficient of 1. All bad wafers representing the case continuously pass through the machine, and there is no good wafer during the abnormal period (bad_period).
Fig. 3 illustrates a correlation representation of a particular tool (i.e., HFXXR 02) for a particular process Step (i.e., step 31). Wherein, the abscissa is the production time sequence of Step31, and the ordinate is the state of the wafer (e.g. 0 or 1 mark of good or bad wafer). As can be seen in connection with fig. 3, all bad wafers continuously pass HFXXR02 for a certain period of time, during which there is no good wafer.
By adopting the embodiment, the correlation degree is processed by combining the factor of the continuity of the wafer abnormality on the basis of calculating the correlation degree, so that the finally obtained correlation coefficient can more accurately represent the commonality correlation of the machine/machine chamber/process step and the target test item. Specifically, the correlation coefficient is obtained after normalization by using a base value, and the base value is determined according to the abnormal wafer continuous value, so that the correlation coefficient can reflect the difference of the correlation of different dimensions to be analyzed on the continuity of wafer abnormality, such as the influence of intermittent abnormality and continuous abnormality in a period of time on the correlation calculation result. Thus, the accuracy of the commonality correlation analysis can be greatly improved.
Furthermore, the present embodiment can realize quantitative analysis of the commonality correlation, and precisely quantify the correlation between the process steps or wafer manufacturing tool units and the target test parameters.
Further, the embodiment can solve the problem of a single machine or a certain process step which cannot be solved by the prior art. Specifically, since the analysis of the commonality correlation in this embodiment is not dependent on the differences between the multiple tools of the same process step, but rather quantifies the correlation coefficient of each tool/chamber of each process step, the analysis of the commonality correlation can still be accurately performed when a particular process step is performed by a single wafer fabrication tool unit. Further, since the analysis dimension of the present embodiment is no longer limited to the differences between the machines, the analysis of the commonality correlation can still be accurately performed when problems occur in the process steps themselves.
In a variation of the present embodiment, the process of calculating the correlation in step S101 may include: and counting the total number of abnormal wafers associated with the dimension to be analyzed in the wafer data to obtain the correlation degree.
That is, in the present modification, the process of determining the abnormal wafer duration vector and the base vector is replaced with the abnormal wafer count process, which can also obtain the correlation operation result without considering the wafer abnormality duration once. For example, the total number of abnormal wafers passing through the Step1 HFXXX01 machine in table 2 is 6, and the total number of abnormal wafers passing through the Step1 HFXXX02 machine in table 3 is 4, which are consistent with the correlation results calculated based on the vector similarity in the foregoing embodiments.
Further, on the basis of the correlation obtained by counting, step S102 is continued to determine a base value based on the abnormal wafer duration value, and the correlation is normalized based on the base value. The final obtained correlation coefficient can embody the persistence factor.
In a variation of this embodiment, the process of calculating the correlation in step S101 may also use an existing ANOVA algorithm. Therefore, on the basis of calculating the correlation based on the existing ANOVA algorithm, the variation continues to execute the step S102 and the step S103, and the correlation coefficient used for sorting in the step S104 can reflect the difference of the correlation of different dimensions to be analyzed on the persistence of the wafer abnormality.
Fig. 4 is a schematic structural diagram of a commonality analysis apparatus 4 of a wafer manufacturing tool unit according to an embodiment of the present invention. It will be appreciated by those skilled in the art that the commonality analyzing apparatus 4 of the wafer manufacturing tool unit according to the present embodiment may be used to implement the method solutions described in the embodiments described above with reference to fig. 1 to 3.
Specifically, referring to fig. 4, the commonality analysis apparatus 4 of the wafer manufacturing tool unit according to the present embodiment may include: a first processing module 41, configured to calculate a correlation degree based on wafer data, where the correlation degree is used to characterize a correlation degree between a dimension to be analyzed and a target test item, and the wafer data is data of a related wafer obtained based on a process step of a production flow; a determining module 42, configured to determine a base value at least according to an abnormal wafer duration value associated with the dimension to be analyzed, where the abnormal wafer duration value is a number of wafers from a first abnormal wafer to a last abnormal wafer in the wafer data in order of production time; a second processing module 43, configured to normalize the correlation based on the base value to obtain a correlation coefficient; a third processing module 44, configured to obtain an analysis result of the dimension to be analyzed based on the value of the correlation coefficient; wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
For more details of the working principle and the working manner of the commonality analysis device 4 of the wafer manufacturing machine unit, reference may be made to the related descriptions in fig. 1 to 3, and the description thereof will not be repeated here.
Further, the embodiment of the invention also discloses a computer readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method technical scheme of any one of the embodiments shown in the above fig. 1 to 3 is executed. Preferably, the computer-readable storage medium may include a computer-readable storage medium such as a non-volatile (non-volatile) memory or a non-transitory (non-transient) memory. The computer readable storage medium may include ROM, RAM, magnetic or optical disks, and the like.
Further, the embodiment of the invention also discloses a terminal, which comprises a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor executes the technical scheme of the method in any one of the embodiments shown in the figures 1 to 3 when running the computer program. Specifically, the terminal may be a computer, a server, or the like.
Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention, and the scope of the invention should be assessed accordingly to that of the appended claims.
Claims (11)
1. A method for analyzing commonality of wafer fabrication tool units, comprising:
calculating correlation based on wafer data, wherein the correlation is used for representing the correlation degree of the dimension to be analyzed and the target test item, and the wafer data is data of related wafers obtained based on the process steps of the production flow;
determining a basic value at least according to the abnormal wafer continuous value associated with the dimension to be analyzed, wherein the abnormal wafer continuous value is the number of wafers from the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence;
carrying out normalization processing on the correlation degree based on the base value to obtain a correlation coefficient;
based on the value of the correlation coefficient, obtaining an analysis result of the dimension to be analyzed;
wherein the dimension to be analyzed is at least selected from: the wafer manufacturing machine unit is used for performing the process steps.
2. The method of claim 1, wherein calculating a correlation based on wafer data comprises:
statistically obtaining abnormal wafer continuous values associated with the dimension to be analyzed from the wafer data;
determining the length of a base vector at least according to the abnormal wafer continuous value, wherein the base vector comprises elements for representing that the state of the wafer is abnormal;
Extracting an abnormal wafer continuous vector associated with the dimension to be analyzed from the wafer data, wherein the abnormal wafer continuous vector comprises elements at least used for representing the states of all wafers from the first abnormal wafer to the last abnormal wafer which are sequenced according to the production time in the wafer data, and the length of the abnormal wafer continuous vector is consistent with the length of the base vector;
and calculating the similarity between the abnormal wafer continuous vector and the base vector to obtain the correlation.
3. The method of commonality analysis of claim 2, wherein said determining a length of a basis vector from at least the abnormal wafer duration value comprises:
counting the total number of abnormal wafers associated with the process steps in the wafer data;
and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the length of the base vector.
4. The method of claim 3, wherein extracting the abnormal wafer persistence vector associated with the dimension to be analyzed from the wafer data comprises:
if the number of the first abnormal wafer to the last abnormal wafer in the wafer data according to the production time sequence is smaller than the length of the base vector, filling preset elements in the abnormal wafer continuous vector so that the length of the abnormal wafer continuous vector is consistent with the length of the base vector.
5. The method of claim 1, wherein calculating a correlation based on wafer data comprises:
and counting the total number of abnormal wafers associated with the dimension to be analyzed in the wafer data to obtain the correlation degree.
6. The method of claim 1, wherein determining a base value based at least on the abnormal wafer persistence value associated with the dimension to be analyzed comprises:
counting the total number of abnormal wafers associated with the process steps in the wafer data;
and determining the larger value of the abnormal wafer duration value and the total number of the abnormal wafers as the base value.
7. The method of claim 1, wherein the wafer data comprises a plurality of records, wherein each record comprises an identification of a wafer and a test value of the wafer at the target test item, the method further comprising:
for each record, marking the state of the wafer according to the preset threshold value of the target test item and the test value, wherein the state comprises abnormal and normal states.
8. The method according to claim 1, characterized in that for correlation coefficients of the same process step in a plurality of dimensions to be analyzed, the correlation coefficient with the largest value is taken as the correlation coefficient of the process step.
9. The method of claim 1, wherein the wafer fabrication tool unit comprises a wafer fabrication tool or a chamber of the wafer fabrication tool.
10. The method of claim 1, wherein obtaining the analysis result of the dimension to be analyzed based on the value of the correlation coefficient comprises:
sequencing the dimension to be analyzed of all the process steps according to the numerical value of the correlation coefficient to obtain an analysis result; or alternatively
And sequencing the dimensions to be analyzed of all the process steps according to the numerical value of the correlation coefficient and a preset coefficient to obtain an analysis result, wherein the preset coefficient is associated with the dimensions to be analyzed.
11. A terminal comprising a memory and a processor, the memory having stored thereon a computer program executable on the processor, characterized in that the processor executes the steps of the method according to any of claims 1 to 10 when the computer program is executed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211591455.6A CN116230586B (en) | 2022-12-12 | 2022-12-12 | Commonality analysis method and terminal of wafer manufacturing machine unit |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211591455.6A CN116230586B (en) | 2022-12-12 | 2022-12-12 | Commonality analysis method and terminal of wafer manufacturing machine unit |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116230586A true CN116230586A (en) | 2023-06-06 |
CN116230586B CN116230586B (en) | 2024-01-02 |
Family
ID=86573782
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211591455.6A Active CN116230586B (en) | 2022-12-12 | 2022-12-12 | Commonality analysis method and terminal of wafer manufacturing machine unit |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116230586B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117272122A (en) * | 2023-11-20 | 2023-12-22 | 全芯智造技术有限公司 | Wafer anomaly commonality analysis method and device, readable storage medium and terminal |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014006807A1 (en) * | 2012-07-03 | 2014-01-09 | シャープ株式会社 | Graph drawing device, graph drawing method, process management system, process management method, and readable storage medium |
CN105990178A (en) * | 2015-02-25 | 2016-10-05 | 中芯国际集成电路制造(上海)有限公司 | Abnormal platform detection method and detection device |
TW202139114A (en) * | 2020-03-31 | 2021-10-16 | 世界先進積體電路股份有限公司 | System and method for determining cause of abnormality in semiconductor manufacturing processes |
US20210366750A1 (en) * | 2019-02-12 | 2021-11-25 | Spp Technologies Co., Ltd. | Abnormality detection apparatus |
CN114266223A (en) * | 2021-12-24 | 2022-04-01 | 上海集成电路装备材料产业创新中心有限公司 | Fault determination method, device, equipment and computer readable storage medium of machine station |
CN114333254A (en) * | 2021-12-27 | 2022-04-12 | 中电九天智能科技有限公司 | Monitoring method and system based on abnormal conditions of continuous batches of semiconductor products |
-
2022
- 2022-12-12 CN CN202211591455.6A patent/CN116230586B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014006807A1 (en) * | 2012-07-03 | 2014-01-09 | シャープ株式会社 | Graph drawing device, graph drawing method, process management system, process management method, and readable storage medium |
CN105990178A (en) * | 2015-02-25 | 2016-10-05 | 中芯国际集成电路制造(上海)有限公司 | Abnormal platform detection method and detection device |
US20210366750A1 (en) * | 2019-02-12 | 2021-11-25 | Spp Technologies Co., Ltd. | Abnormality detection apparatus |
TW202139114A (en) * | 2020-03-31 | 2021-10-16 | 世界先進積體電路股份有限公司 | System and method for determining cause of abnormality in semiconductor manufacturing processes |
CN114266223A (en) * | 2021-12-24 | 2022-04-01 | 上海集成电路装备材料产业创新中心有限公司 | Fault determination method, device, equipment and computer readable storage medium of machine station |
CN114333254A (en) * | 2021-12-27 | 2022-04-12 | 中电九天智能科技有限公司 | Monitoring method and system based on abnormal conditions of continuous batches of semiconductor products |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117272122A (en) * | 2023-11-20 | 2023-12-22 | 全芯智造技术有限公司 | Wafer anomaly commonality analysis method and device, readable storage medium and terminal |
CN117272122B (en) * | 2023-11-20 | 2024-04-02 | 全芯智造技术有限公司 | Wafer anomaly commonality analysis method and device, readable storage medium and terminal |
Also Published As
Publication number | Publication date |
---|---|
CN116230586B (en) | 2024-01-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR102384189B1 (en) | Method and apparatus for predicting semiconductor failure time based on machine learning | |
CN116230586B (en) | Commonality analysis method and terminal of wafer manufacturing machine unit | |
CN108491991B (en) | Constraint condition analysis system and method based on industrial big data product construction period | |
US20070094196A1 (en) | Manufacture data analysis method and manufacture data analyzer apparatus | |
CN116113942A (en) | Predicting equipment failure modes based on process traces | |
CN113009823A (en) | Method and apparatus for controlling a machine | |
JP7354421B2 (en) | Error factor estimation device and estimation method | |
CN111401420B (en) | Abnormal data clustering method and device for wafer test, electronic equipment and medium | |
WO2020166236A1 (en) | Work efficiency evaluating method, work efficiency evaluating device, and program | |
US8649990B2 (en) | Method for detecting variance in semiconductor processes | |
CN113672506B (en) | Dynamic proportion test case sorting and selecting method and system based on machine learning | |
US10656204B2 (en) | Failure detection for wire bonding in semiconductors | |
Huang et al. | Towards smarter diagnosis: A learning-based diagnostic outcome previewer | |
JP2019032671A (en) | Cause estimation method and program | |
KR20210065751A (en) | System and method for estimating a missing value | |
CN112181792A (en) | Method, system and related assembly for abnormal marking of time sequence data | |
Bodhe et al. | Diagnostic Fail Data Minimization Using an $ N $-Cover Algorithm | |
Anaya et al. | Yield improvement using advanced data analytics | |
US20100010763A1 (en) | Method for detecting variance in semiconductor processes | |
Nagy et al. | An industrial application using process mining to reduce the number of faulty products | |
JP4866520B2 (en) | Data analysis method, data analysis program, and data analysis apparatus | |
CN114580982B (en) | Method, device and equipment for evaluating data quality of industrial equipment | |
Kerdprasop et al. | Performance analysis of complex manufacturing process with sequence data mining technique | |
Hsu et al. | Main branch decision tree algorithm for yield enhancement with class imbalance | |
Kim et al. | A sensor data mining process for identifying root causes associated with low yield in semiconductor manufacturing |
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 |