Data analysis apparatus, data analysis system, and data analysis method
The data analysis device optimizes component combinations in semiconductor manufacturing equipment by predicting performance and adjusting parts, addressing performance variations and reducing costs and time.
Patent Information
- Application Number
- JP2024112064
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-07-11
- Publication Date
- 2026-01-23
AI Technical Summary
Existing semiconductor manufacturing equipment technologies fail to address performance variations between devices due to component differences, necessitating strict tolerance control, which is time-consuming and costly.
A data analysis device that predicts equipment performance by analyzing component characteristics, creating a prediction model to optimize component combinations, and adjusts parts to ensure performance within predetermined ranges.
Reduces performance variations between devices by optimizing component combinations, reducing costs and time required for selection, without stringent tolerance management.
Smart Images
Figure 2026011457000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to a data analysis device, a data analysis system, and a data analysis method. [Background technology]
[0002] In the field of semiconductor manufacturing, it is known that even if multiple semiconductor manufacturing devices with common configurations are operated in the same way, there are differences in the occurrence of defects and operating rates between semiconductor manufacturing devices (individual differences).
[0003] For example, Patent Document 1 proposes a technology for comparing processing states and results of processing by multiple processing devices. Patent Document 1 discloses the following content: "A semiconductor manufacturing method characterized by comprising a system for managing multiple processing devices when processing semiconductors using multiple processing devices, inspecting the operating conditions and processing states of the multiple processing devices, inspecting the semiconductors after processing is completed, accumulating and processing desired data, comparing the processing states and results of the multiple processing devices, and managing the multiple processing devices based on this." [Prior art documents] [Patent documents]
[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-134376 Summary of the Invention [Problem to be solved by the invention]
[0005] Although Patent Document 1 focuses on data processing of semiconductor manufacturing equipment in operation, it does not mention data processing in a state before the equipment is put into operation (for example, during quality assurance inspection or assembly inspection). Therefore, Patent Document 1 does not consider suppressing performance variations between semiconductor manufacturing equipment before the equipment is put into operation. Furthermore, semiconductor manufacturing equipment is constructed by combining multiple components with different functions, and variations between the components can cause differences between the equipment. Therefore, if we were to try to suppress variations between the components during assembly to prevent variations in performance between the equipment, strict control of the tolerances of the component dimensions and the allowable ranges of the setting values would be required, which would create another problem: it would take time and money to select the components. Therefore, an object of the present invention is to provide a technique that makes it possible to suppress variations in performance between devices even when there are variations between parts. [Means for solving the problem]
[0006] In order to solve the above-mentioned problems, one representative data analysis device of the present invention is a data analysis device for selecting components that constitute semiconductor manufacturing equipment, and includes: a data acquisition unit that acquires equipment data that indicates the performance of the semiconductor manufacturing equipment and component data that indicates the characteristics of the components that constitute the semiconductor manufacturing equipment; a data processing unit that uses the equipment data and the component data to create a prediction model that predicts equipment data of a target device that is the semiconductor manufacturing equipment to be analyzed; a data determination unit that inputs component data of the components that constitute the target device into the prediction model, predicts the equipment data of the target device, and determines whether the predicted equipment data is within a predetermined range; and a data selection unit that changes the combination of components that constitute the target device if it is determined that the predicted equipment data is not within the predetermined range. [Effects of the Invention]
[0007] According to the present invention, even when there is variation between parts, it is possible to suppress variation in performance between devices. Problems, configurations, and effects other than those described above will become apparent from the following description of the preferred embodiment of the invention. [Brief explanation of the drawings]
[0008] [Figure 1]FIG. 1 is a diagram illustrating an example of the configuration of a data analysis system. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of the inspection computer. [Figure 3] FIG. 3 is a diagram illustrating an example of data stored in the server system. [Figure 4] FIG. 4 is a diagram illustrating an example of the configuration of the analysis computer. [Figure 5] FIG. 5 is a flowchart showing the processing in the analysis computer. [Figure 6] FIG. 6 is a flowchart showing the processing relating to device data. [Figure 7] FIG. 7 is a diagram illustrating an example of the configuration of the analysis computer. DETAILED DESCRIPTION OF THE INVENTION
[0009] Hereinafter, examples will be described with reference to the drawings. However, the present invention is not limited to these examples. In the drawings, the same parts are denoted by the same reference numerals.
[0010] In this disclosure, "equipment data" refers to data that indicates the performance of semiconductor manufacturing equipment (in this disclosure, this refers to a plasma etching equipment). For example, when a plasma etching equipment is shipped as a product, the equipment data can be used as inspection data that is attached to the product to guarantee its performance, and therefore is also referred to as "shipping inspection data." "Component data" refers to data that indicates the characteristics of the components that make up the plasma etching equipment. This component data is sometimes created by the manufacturer of the plasma etching equipment components and attached together with the inspection data of the plasma etching equipment that is treated as a product, so it is also called "manufacturer shipping inspection data." [Example]
[0011] (Configuration of semiconductor manufacturing equipment) A data analysis system 1 for semiconductor manufacturing equipment will be described using a plasma etching equipment, which is one type of semiconductor manufacturing equipment, as an example. FIG. 1 is a diagram showing an example of the configuration of data analysis system 1. Data analysis system 1 includes plasma etching equipment 100, inspection computer 111, server system 112, and analysis computer 301. Plasma etching equipment 100 includes control computer 110 connected via signal lines to each of the control devices, namely, flow meter 104, pressure sensor 105, turbomolecular pump 106, conductance valve (VV) 107, high-frequency power supply 108, and matching box 109.
[0012] The processing chamber 103 has a lower electrode 102 disposed therein and on which a substrate 101 to be processed is placed. A high-frequency power supply 108 is a power supply that applies high-frequency power to the lower electrode 102, and is connected to the lower electrode 102 via a matcher 109.
[0013] The processing chamber 103 is also provided with a flow meter 104, a pressure sensor 105, and a conductance valve 107. The flow meter 104 controls the amount of process gas supplied into the processing chamber 103, and is implemented, for example, by a mass flow controller. The pressure sensor 105 is a sensor that measures the gas pressure inside the processing chamber 103, and is implemented, for example, by a Baratron vacuum gauge. The conductance valve 107 is disposed between the processing chamber 103 and a turbomolecular pump 106. The turbomolecular pump 106 is a pump that exhausts the process gas inside the processing chamber 103. The conductance valve 107 adjusts the amount of exhausted process gas, thereby adjusting the degree of vacuum inside the processing chamber 103.
[0014] The plasma etching apparatus 100 also includes a control computer 110. The control computer 110 is connected to each of a plurality of control devices, namely, a flow meter 104, a pressure sensor 105, a conductance valve 107, a high-frequency power supply 108, and a matching box 109, by signal lines, and controls these control devices. The control computer 110 communicates with an inspection computer 111 that is external to the plasma etching apparatus 100.
[0015] In data analysis system 1, for example, a performance inspection (e.g., quality inspection or assembly inspection) of plasma etching apparatus 100 is performed. Plasma etching apparatus 100 is put into a trial operating state, and data necessary for analysis is acquired. In addition to plasma etching apparatus 100, data analysis system 1 can also include other plasma etching apparatuses having the same configuration. In the following explanation, data processing for plasma etching apparatus 100 will be described, but similar data processing can also be performed for other plasma etching apparatuses.
[0016] (Considerations regarding differences between devices) As described above, plasma etching apparatus 100 is configured as a single apparatus by combining multiple components with different functions. If multiple plasma etching apparatuses are used, variations in dimensions and settings may occur even among common components (hereinafter simply referred to as "variations between components"), which may result in differences (variations) in etching performance between the apparatuses.
[0017] In the past, in order to suppress (reduce) variations in etching performance, variations between components were reduced to reduce differences between plasma etching equipment (hereinafter simply referred to as "equipment-to-equipment differences"), thereby achieving uniform etching. However, in recent years, semiconductor devices have undergone increasingly miniaturized and complex processing, and in addition, the need to meet the demands of mass production of semiconductor devices has created a demand for further reductions in differences between plasma etching equipment. However, when employing a method to reduce variations between components to reduce differences between equipment, strict control of component dimensional tolerances and allowable ranges of setting values is required, which creates another problem: component selection is time-consuming and expensive.
[0018] Therefore, this disclosure presents a method for reducing the differences between devices by optimizing the combination of parts even when there is variation between the parts. This disclosure uses device data of a plasma etching device as an index of device performance, and optimizes the combination of parts so as to reduce the variation between the device data.
[0019] (control computer) The control computer 110 receives process data PD, which is data including multiple items that are obtained from each control device and indicate the operating status of the plasma etching apparatus 100. 100 , is acquired as digital data. Process data PD 100 The data included in the table includes, for example, the gas flow rate obtained from the flow meter 104, the gas pressure obtained from the pressure sensor 105, the valve opening of the conductance valve 107, or the process parameters of the high frequency power source 108 and the matching box 109 in plasma processing.
[0020] (Testing computer) The inspection computer 111 is a device that performs and manages performance inspections of the plasma etching apparatus 100. The inspection computer 111 adds PD to the process data. 100The performance of the plasma etching apparatus 100 is inspected based on the data, and the apparatus data SD 100 Specifically, the inspection computer 111 has a data recording format (for example, a PC form) in which the inspection items for the performance inspection of the plasma etching apparatus 100 and the pass criteria for each inspection item are set, and the process data PD output from the control computer 110 is calculated. 100 and extract the necessary data according to the inspection item. In other words, the equipment data SD 100 is calculated based on data including multiple items.
[0021] 2 is a diagram showing an example of the configuration of the testing computer 111. The testing computer 111 includes a processor 1111, a memory 1112, a recording medium 1113, and a communication device 1114. The processor 1111, the memory 1112, the recording medium 1113, and the communication device 1114 are connected to one another via a bus 1110. The processor 1111 controls the memory 1112, the recording medium 1113, and the communication device 1114 via the bus 1110, and also performs necessary calculations.
[0022] The memory 1112 includes processing instructions that cause the processor 1111 to perform predetermined processing. For example, the processor 1111 executes the processing instructions to generate process data PD 100 The test computer 111 can function as a test means for testing performance based on the test results and as a calculation means for calculating device data. The recording medium 1113 stores data in accordance with the data recording format. The communication device 1114 transmits and receives data between the test computer 111 and the control computer 110, and also transmits and receives data to and from the server system 112, which will be described later.
[0023] (Server system) The server system 112 receives the device data SD calculated by the test computer 111. 100 The server system 112 also stores component data BD indicating the characteristics of the components that make up the semiconductor manufacturing equipment 100.100 Save the part data BD 100 This can be obtained, for example, from the manufacturer of the parts of the semiconductor manufacturing equipment 100. The server system 112 stores the data in a format that can be referenced by the analysis computer 301, which will be described later.
[0024] 3A and 3B are diagrams showing an example of data stored in the server system 112. FIG. 3A shows equipment data SD, and FIG. 3B shows part data BD. Note that data for each plasma etching equipment is indicated with a subscript, and the entire data stored in the server system 112 is indicated without the subscript. For example, equipment data SD for plasma etching equipment A is A , plasma etching equipment B equipment data SD B , plasma etching equipment C equipment data SD C The device data SD of the server system 112 is shown as follows: A From SD C The same applies to the part data BD.
[0025] In FIG. 3(a), the apparatus data SD of the plasma etching apparatus 100 100 is xx, and the equipment data SD of plasma etching equipment A A is aa, and the equipment data SD of plasma etching equipment B B is bb, and the equipment data SD of the plasma etching equipment C C is cc. Device data SD 100 is a plurality of output values acquired from the plasma etching apparatus 100, a parameter value set in the plasma etching apparatus 100, or an index calculated from the plurality of output values and the parameter value. 100The calculation method can be set as appropriate. For example, a specific output value or parameter of the plasma etching apparatus 100 may be used as the apparatus data. It may also be a numerical value based on a semiconductor structure formed on the substrate 101 to be processed by performing a performance inspection process on the substrate 101 to be processed. Furthermore, although the case where it is expressed as a single numerical value has been shown, the apparatus data may also be data including multiple items.
[0026] Figure 3(a) also shows the judgment results for the equipment data variation, which can be compared with the pass criteria set in the data recording format for each inspection item, and if the pass criteria are met, the result is OK (equipment data is not within the tolerance range), and if the pass criteria cannot be met, the result is NG (equipment data is within the tolerance range). Figure 3(a) shows that the judgment results for plasma etching equipment A and D are OK, and the judgment results for plasma etching equipment B and C are NG.
[0027] In FIG. 3(b), the part data BD is the part data BD of the plasma etching equipment 100. 100 The case where the part data BD is included is shown. 100 indicates data on the flow meter, pressure sensor, conductance valve, high-frequency power supply, and matching box, which are components that make up the plasma etching apparatus 100. The flow meter, pressure sensor, conductance valve, and matching box each include the items "manufacturer," "dimensions," and "evaluation value." In addition, the high-frequency power supply includes "output power." The item "manufacturer" indicates the manufacturer or supplier of the component, and the item "dimensions" indicates the size of a specific location. The item "output power" indicates the output power value. The item "evaluation value" evaluates the value of each component item and indicates the evaluation result numerically. Note that each item is an example, and the present disclosure is not limited to this case. For example, while the case where the number of components is 5 is shown, component data BD may be generated for n components (n is an integer) selected by the user from among the components that make up the plasma etching apparatus 100. 100Alternatively, items for each part may be set in the data recording format, and the items set in the data recording format may be adopted. As for the method of calculating the evaluation value, the data recording format may also set conditions necessary for calculation, such as conditions for quantifying the evaluation value and weights between items.
[0028] (Analysis computer (data analysis device)) 4 is a diagram showing an example of the configuration of analytical computer 301. The analytical computer is a device for selecting components that constitute a plasma etching apparatus. Analytical computer 301 includes processor 3011, memory 3012, recording medium (data storage unit) 3013, communication device 3014, and user interface unit 3015. Processor 3011, memory 3012, recording medium 3013, communication device 3014, and user interface unit 3015 are connected to one another via bus 3010. Processor 3011 controls memory 3012, recording medium 3013, communication device 3014, and user interface unit 3015 via bus 3010, and performs necessary calculations.
[0029] Memory 3012 contains processing instructions that cause processor 3011 to execute predetermined processes. Recording medium 3013 stores data acquired from server system 112 and also stores data calculated by processor 3011. Communication device 3014 transmits and receives data between analytical computer 301 and server system 112. User interface unit 3015 accepts user instructions and presents information such as analysis results to the user.
[0030] The processor 3011 functions as a data acquisition unit (data acquisition means) 302, a data processing unit (data processing means, model generation unit (model generation means)) 303, a data determination unit (data determination means) 304, and a data selection unit (data selection means) 305. The data acquisition unit 302 acquires equipment data SD 100 and component data BD indicating the characteristics of the components that make up the semiconductor manufacturing equipment 100. 100The acquired data is stored in the recording medium 3013. The data processing unit 303 acquires the device data SD 100 and parts data BD 100 The equipment data SD of the target equipment, which is the semiconductor manufacturing equipment to be analyzed, is used. t In creating this model, processes such as difference, average, standard deviation, variance, minimum, maximum, correlation such as correlation analysis and regression analysis, statistical analysis process, and machine learning process are performed. t is input into the prediction model to predict the equipment data of the target equipment, and the predicted equipment data SD tp The data selection unit 305 determines whether the predicted device data SD tp If it is determined that the difference is not within the predetermined range, the combination of parts constituting the target device is changed. Each function of the processor 3011 will be described later.
[0031] (Analysis processing) FIG. 5 is a flowchart showing the processing in analytical computer 301.
[0032] Step 401: The device data SD is stored in the recording medium 3013. The data acquisition unit 302 acquires the device data SD stored in the server system 112 and stores it in the recording medium 3013.
[0033] Step 402: The part data BD is stored on the recording medium 3013. For example, the user may store the part data BD in advance in the server system 112. Then, the data acquisition unit 302 acquires the part data BD stored in the server system 112 and stores it on the recording medium 3013.
[0034] Step 403: The correlation between the component data DB and the equipment data SD is calculated. The data processing unit 303 performs statistical analysis using the component data DB and the equipment data SD. For example, the component data DB and the equipment data SD are collected for multiple plasma etching equipment, and statistical analysis is performed to extract the mean μ, standard deviation σ, variance V, minimum value, maximum value, correlation coefficient, etc., to understand the statistical characteristics of the component data DB and the equipment data SD.
[0035] Step 404: A prediction model for predicting the equipment data is created. The data processing unit 303 creates the prediction model by, for example, performing regression analysis. For example, the equipment data is used as the objective variable y, and the component data BD (evaluation value for each component) is used as the explanatory variable x1, x2, x3, ... x n , the regression coefficients are β1, β2, β3…β n In this case, the following formula (1) is derived as a prediction model. Note that, although this disclosure describes a case where a prediction model is generated by performing regression analysis, which is a statistical analysis method, as a method for generating a prediction model, the method for generating a prediction model is not limited to statistical analysis. For example, a prediction model may be generated by performing machine learning using the device data SD and the part data BD as training data.
number
[0036] Step 405: The part data BD of the plasma etching equipment to be analyzed (hereinafter also referred to as the "target equipment") t are input to the prediction model (Equation (1)). The plasma etching equipment to be analyzed does not necessarily have to be in an assembled state. For example, for a plasma etching equipment before assembly, data on parts to be used in assembly may be stored as parts data BD t is substituted into equation (1).
[0037] Step 406: Calculation of equipment data of the analysis target The data determination unit 304 inputs the component data input in step S405 into equation (1) to predict equipment data of the plasma etching equipment to be analyzed.
[0038] Step 407: Predicted device data SD tp The data determination unit 304 determines the performance of the target device based on the predicted device data SD tp is included in a predetermined range. For example, the predetermined range is set to ±3σ using the standard deviation σ of the device data SD extracted in step 403. If the predicted device data exceeds ±3σ, the data judgment unit 304 judges that the performance of the target device is poor (NG). Also, if the predicted device data SD tp If is within the range of ±3σ, the data determination unit 304 determines that the performance of the target device is good (OK).
[0039] Step 408: If the result of the determination in step 407 is NG, an instruction to change the combination of parts is issued. The data selection unit 305 selects the predicted device data SD tp and the target device data. The change instruction is sent from the data selection unit 305 to the data processing unit 303, which extracts parts having evaluation values from the part data BD. Also, it is possible to use the average value μ of the device data SD extracted in step 402, for example, as the target device data. With the combination of parts changed, the process returns to step 406 again.
[0040] A change in the combination of parts is determined based on the regression coefficients calculated from the regression analysis. The data processing unit 303 extracts candidate parts for constituting the target device using the regression coefficients of the prediction model, and the data selection unit 305 selects a new combination of parts from the candidate parts. Since parts with large regression coefficients have a significant impact on the variability of device data, a combination of parts corresponding to the regression coefficients is recalculated using the prediction model in step S406, and the device data for the target device is calculated. The determination result in step S407 and an instruction to change the combination in step S408 may be notified to the user via the user interface unit 3015. The user can also change the combination of parts based on the determination result.
[0041] Step 409: If the result of the determination in step 407 is OK, a shipping inspection is carried out. t The plasma etching equipment is assembled using the computer 111, and the assembled plasma etching equipment is inspected before shipping. The inspection computer 111 inspects the assembled plasma etching equipment before shipping, and stores the equipment data and parts data in the server system 112.
[0042] (Actions and Effects) In this way, analytical computer 301 uses component data BD and equipment data SD stored on recording medium 3013 to create a prediction model for predicting equipment data in data processing unit 303. Then, analytical computer 301 predicts equipment data for a hypothetical pre-assembly plasma etching equipment using the prediction model. If analytical computer 301 determines that performance is poor based on the predicted equipment data, it selects a combination of components that will bring the equipment data closer to the target equipment data and issues an instruction to change the components.
[0043] By using analytical computer 301 to change the combination of parts before assembling the plasma etching equipment, it was possible to suppress the variation in performance between plasma etching equipment, even when the tolerance range of some of the components of the plasma etching equipment was expanded by 2% to 5%. As a result, parts that previously had large tolerances and could not be installed in equipment can now be used, reducing the cost and time required for part selection.
[0044] As described above, the present disclosure can reduce performance variations between equipment even when there are variations between components. By selecting a combination of components that reduces the variations between equipment data, it is possible to reduce differences between equipment without requiring strict management of the dimensional tolerances of each component or the allowable range of setting values. Furthermore, because there is no need to strictly manage variations between components, it is possible to reduce the cost and time required to select components that make up semiconductor manufacturing equipment. [Example]
[0045] Example 2 differs from Example 1 in that equipment data is repeatedly acquired from one plasma etching equipment. Fig. 6 is a flowchart showing processing related to equipment data. In the following description, components that are the same as or equivalent to those in Example 1 above are given the same reference numerals, and their description will be simplified or omitted.
[0046] As the first process P1, data is collected. The inspection computer 111 collects process data PD 100 The inspection computer 111 collects the process data PD from each control device of the plasma etching apparatus 100 via the control computer 110. 100 The test computer 111 acquires the process data PD 100 A performance test was carried out using the process data PD 100 and device data SD calculated as a result of performance testing 100 is stored in the recording medium 1113 in accordance with the data recording format.
[0047] Next, as the second process P2, data is stored. The inspection computer 111 stores the data in the data recording format held in the recording medium 1113 in the server system 112. The server system 112 records the data in the data recording format in a format that can be accessed by the analysis computer 301, and stores it as a database. The server system 112 also stores the parts data BD 100 is also stored together with other data.
[0048] Next, as the third process P3, the data is plotted in time series. The data determination unit 304 of the analytical computer 301 selects the device data SD 100 The data determination unit 304 extracts the device data SD 100 The time-series fluctuations of the plasma etching apparatus 100 are monitored to perform trend management.
[0049] For the equipment data for which time-series fluctuations have been confirmed, the data selection unit 305 identifies the influence of each component on the fluctuations in the inspection data, and selects components and changes the combination of components so as to suppress the fluctuations in the inspection data. In this way, the analysis computer 301 can suppress time-series fluctuations (variations) in the inspection data. [Example]
[0050] In the first embodiment, the performance of semiconductor manufacturing equipment was evaluated statistically using previously acquired equipment data SD and component data BD. On the other hand, the third embodiment differs from the first embodiment in that the analytical computer includes an instrument difference analysis unit that analyzes instrument differences between the equipment being analyzed. FIG. 7 is a diagram showing an example of the configuration of analytical computer 301A. In the following description, components that are the same as or equivalent to those in the first embodiment are designated by the same reference numerals, and their description will be simplified or omitted.
[0051] The analytical computer 301A has a machine difference determination unit 306. In a method of analyzing machine differences when there are two target devices, the data determination unit 304 first inputs the device data of the other target device, which is a semiconductor manufacturing device different from the target device, into a prediction model, and predicts the device data of the other target device. Next, the machine difference determination unit 306 calculates the device data SD of the target device. tp and the device data of the other target devices, it is determined whether the machine difference between the target device and the other target devices is normal. Whether or not it is normal can be determined using, for example, the standard deviation σ extracted by the data processing unit 303.
[0052] In the third embodiment, it becomes possible to evaluate the performance variation between target devices. For example, when a manufacturer that supplies parts is changed (a supplier change), it becomes possible to evaluate the parts of the new manufacturer by checking the machine differences between the semiconductor manufacturing equipment using parts from the previous manufacturer and the semiconductor manufacturing equipment using parts from the new manufacturer.
[0053] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the gist of the present invention.
[0054] The following are examples of possible embodiments of the present invention, but the present invention is not limited to these. (Aspect 1) A data analysis device for selecting components constituting a semiconductor manufacturing device, a data acquisition unit that acquires equipment data indicating the performance of a semiconductor manufacturing equipment and parts data indicating the characteristics of parts that constitute the semiconductor manufacturing equipment; a data processing unit that uses the equipment data and the component data to create a prediction model that predicts equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; a data determination unit that inputs part data of parts constituting the target device into the prediction model, predicts device data of the target device, and determines whether the predicted device data is within a predetermined range; a data selection unit that changes a combination of parts that make up the target device when it is determined that the predicted device data does not fall within the predetermined range. (Aspect 2) The components constituting the semiconductor manufacturing equipment include a plurality of control devices, the semiconductor manufacturing equipment has a control computer connected to the plurality of control devices, 2. The data analysis device according to aspect 1, wherein the device data is calculated based on data relating to the plurality of control devices output from the control computer. (Aspect 3) 3. The data analysis device according to aspect 1 or 2, wherein the device data is calculated based on data including a plurality of items. (Aspect 4) the data processing unit performs a regression analysis to create the prediction model; 4. The data analysis device according to any one of aspects 1 to 3, wherein the data selection unit selects components that configure the target device using a regression coefficient of the prediction model. (Aspect 5) The data analysis device further includes an instrument difference determination unit, the data determination unit inputs equipment data of another target device, which is a semiconductor manufacturing device different from the target device, into the prediction model, and predicts the equipment data of the other target device; The data analysis device of any one of aspects 1 to 4, wherein the machine difference determination unit determines whether the machine difference between the target device and the other target device is normal based on the device data of the target device and the device data of the other target device. (Aspect 6) the equipment data is a plurality of output values acquired from the semiconductor manufacturing equipment, parameter values set in the semiconductor manufacturing equipment, or an index calculated from the plurality of output values and the parameter values; 6. The data analysis device according to any one of aspects 1 to 5, wherein the component data of the semiconductor manufacturing device includes dimensions of the components. (Aspect 7) The components constituting the semiconductor manufacturing equipment include a power supply, 7. The data analysis device according to claim 1, wherein the component data of the semiconductor manufacturing device includes an output value of the power supply. (Aspect 8) A data analysis system for selecting components constituting semiconductor manufacturing equipment, an inspection computer that calculates equipment data that indicates the performance of the semiconductor manufacturing equipment; a server system that stores the equipment data and parts data that indicate the characteristics of parts that constitute the semiconductor manufacturing equipment; a data analysis device, The data analysis device a data acquisition unit that acquires the device data and the part data from the server system; a data processing unit that uses the equipment data and the component data to create a prediction model that predicts equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; a data determination unit that inputs part data of parts constituting the target device into the prediction model, predicts device data of the target device, and determines whether the predicted device data is within a predetermined range; a data selection unit that changes the combination of parts that make up the target device when it is determined that the predicted device data does not fall within the specified range. (Aspect 9) A data analysis method for selecting components constituting semiconductor manufacturing equipment, comprising: Acquire equipment data indicating the performance of semiconductor manufacturing equipment and parts data indicating the characteristics of parts constituting the semiconductor manufacturing equipment; Using the equipment data and the component data, a prediction model is created to predict equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; inputting part data of parts constituting the target device into the prediction model, predicting device data of the target device, and determining whether the predicted device data is within a predetermined range; A data analysis method comprising the step of changing a combination of parts constituting the target device when it is determined that the predicted device data does not fall within the predetermined range. [Explanation of symbols]
[0055] 100 plasma etching apparatus, 101 substrate to be processed, 102 lower electrode, 103 processing chamber, 104 flow meter, 105 pressure sensor, 106 turbomolecular pump, 107 conductance valve, 108 high frequency power supply, 109 matching box, 110 control computer, 111 inspection computer, 112 Server systems, 301, 301A Analytical computers 302 data acquisition unit, 303 data processing unit, 304 data determination unit, 305 data selection unit, 306 machine difference judgment unit 1110 bus, 1111 processor, 1112 memory, 1113 recording medium, 1114 communication devices; 3010 bus, 3011 processor, 3012 memory, 3013 recording medium, 3014 Communication device, 3015 User interface unit
Claims
1. A data analysis device for selecting components constituting a semiconductor manufacturing device, a data acquisition unit that acquires equipment data indicating the performance of a semiconductor manufacturing equipment and parts data indicating the characteristics of parts that constitute the semiconductor manufacturing equipment; a data processing unit that uses the equipment data and the component data to create a prediction model that predicts equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; a data determination unit that inputs part data of parts constituting the target device into the prediction model, predicts device data of the target device, and determines whether the predicted device data is within a predetermined range; a data selection unit that changes a combination of parts that make up the target device when it is determined that the predicted device data does not fall within the predetermined range.
2. The components constituting the semiconductor manufacturing equipment include a plurality of control devices, the semiconductor manufacturing equipment has a control computer connected to the plurality of control devices, 2. The data analysis device according to claim 1, wherein the device data is calculated based on data relating to the plurality of control devices output from the control computer.
3. The data analysis device according to claim 1 , wherein the device data is calculated based on data including a plurality of items.
4. the data processing unit performs a regression analysis to create the prediction model; 2. The data analysis device according to claim 1, wherein the data selection unit selects components that constitute the target device using regression coefficients of the prediction model.
5. The data analysis device further includes an instrument difference determination unit, the data determination unit inputs equipment data of another target device, which is a semiconductor manufacturing device different from the target device, into the prediction model, and predicts the equipment data of the other target device; The data analysis device according to claim 1 , wherein the machine difference determination unit determines whether the machine difference between the target device and the other target device is normal based on the device data of the target device and the device data of the other target device.
6. the equipment data is a plurality of output values acquired from the semiconductor manufacturing equipment, parameter values set in the semiconductor manufacturing equipment, or an index calculated from the plurality of output values and the parameter values; 4. The data analysis device according to claim 3, wherein the component data of the semiconductor manufacturing device includes dimensions of the components.
7. The components constituting the semiconductor manufacturing equipment include a power supply, The data analysis device according to claim 6 , wherein the component data of the semiconductor manufacturing device includes an output value of the power supply.
8. A data analysis system for selecting components constituting semiconductor manufacturing equipment, an inspection computer that calculates equipment data that indicates the performance of the semiconductor manufacturing equipment; a server system that stores the equipment data and parts data that indicate the characteristics of parts that constitute the semiconductor manufacturing equipment; a data analysis device, The data analysis device a data acquisition unit that acquires the device data and the part data from the server system; a data processing unit that uses the equipment data and the component data to create a prediction model that predicts equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; a data determination unit that inputs part data of parts constituting the target device into the prediction model, predicts device data of the target device, and determines whether the predicted device data is within a predetermined range; a data selection unit that changes the combination of parts that make up the target device when it is determined that the predicted device data does not fall within the specified range.
9. A data analysis method for selecting components constituting semiconductor manufacturing equipment, comprising: Acquire equipment data indicating the performance of semiconductor manufacturing equipment and parts data indicating the characteristics of parts constituting the semiconductor manufacturing equipment; Using the equipment data and the component data, a prediction model is created to predict equipment data of a target equipment, which is a semiconductor manufacturing equipment to be analyzed; inputting part data of parts constituting the target device into the prediction model, predicting device data of the target device, and determining whether the predicted device data is within a predetermined range; A data analysis method comprising the step of changing a combination of parts constituting the target device when it is determined that the predicted device data does not fall within the predetermined range.
Citation Information
Patent Citations
JP134376A