A slicing process for a main control chip

Abstract

The invention discloses a slicing process of a main control chip, comprising S1 preparing a silicon wafer, S2 coating, S3 fixing, S4 cutting, S5 cleaning, S6 grinding and polishing, S7 laser surface scanning and collecting information, S8 data analysis for quality inspection, and S9 subsequent packaging operation; the invention uses a slicing quality control system, which comprises a data acquisition unit, a data cleaning processing unit, a quality inspection and analysis module, a model analysis module, a human-computer interaction unit and a slicing parameter adjustment unit, wherein the quality inspection and analysis module is used to judge the size, surface roughness, damage and cracks of the chip; the model analysis module analyzes and judges the cause of the damage on the chip surface through a learning model, and provides corresponding solutions; by using laser surface scanning technology to collect data, and then using the quality inspection and analysis module to perform multi-faceted automatic quality inspection on the size, roughness, damage and cracks of the chip, the existing quality inspection limitations and the problem of difficulty in discovering hidden defects are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of chip manufacturing, and in particular to a slicing process of a main control chip. Background Art

[0002] The manufacturing process of the main control chip refers to a series of process steps and technologies that convert the designed circuit diagram into an actual usable chip, including chip design, circuit simulation, lithography, etching and slicing; the slicing process of the main control chip is a series of process steps to cut the main control chip from the silicon wafer into individual chips. At the end of the slicing process, the chip is usually inspected by visual inspection, size measurement and other methods to ensure that there are no problems that affect the chip quality during the slicing process. However, the current quality inspection process after the chip slicing process is usually a sampling inspection of a small part of a large batch of products. There are quality inspection limitations, and most of the obvious defects are found through manual visual inspection, but some hidden defects may be difficult to detect, and the automation and consistency of the detection adjustment are insufficient. Therefore, there is an urgent need for a slicing process for the main control chip that can optimize the quality inspection process to ensure the quality of the chip after slicing. Summary of the invention

[0003] In order to solve the problems existing in the slicing process of the main control chip mentioned in the above background technology, the present invention provides a slicing process of the main control chip. The technical solution adopted by the present invention is as follows:

[0004] A slicing process of a main control chip comprises the following steps:

[0005] S1: According to the design requirements, silicon wafers that have completed the preliminary process are used as slicing raw materials. The preliminary process includes cleaning, photolithography, exposure and development;

[0006] S2: Coating a protective film on the surface of the silicon wafer covered with photoresist to protect the chip surface from damage or contamination during the cutting process;

[0007] S3: Place the silicon wafer on the slicer table and fix it in the proper position using a clamping device to ensure that the silicon wafer is precisely aligned with the cutting tool to ensure the accuracy and stability of the cutting;

[0008] S4: Use high-precision cutting tools to cut silicon wafers. The movement of the cutting tools is controlled by the slice quality control system to ensure the accuracy of the cutting line and the size of the chip meets the design requirements;

[0009] S5: Cleaning the cut chips to remove the protective film residues and cutting debris that may remain on the surface;

[0010] S6: Grinding and polishing the sliced ​​chips to make their surfaces smoother and more uniform, which helps to improve the quality and reliability of the chips;

[0011] S7: Setting the parameters of the fixed laser, scanning the surface of the chip moving at a uniform speed on the conveyor belt by using the laser surface scanning technology, and collecting data information on the chip surface;

[0012] S8: Analyze the collected data information using the quality inspection analysis module and the model analysis module to determine whether the chip size, surface roughness, surface damage degree and crack degree meet the specified range;

[0013] S9: Qualified products are sent to the next process for packaging, and unqualified products are subjected to different operations according to the reasons for failure;

[0014] The slice quality control system of S4 includes a data acquisition unit, a data cleaning processing unit, a quality inspection and analysis module, a model analysis module, a human-computer interaction unit and a slice parameter adjustment unit.

[0015] Further, the output end of the data acquisition unit is unidirectionally connected to the input end of the data cleaning processing unit, the output end of the data cleaning processing unit is unidirectionally connected to the input end of the quality inspection and analysis module, the output end of the quality inspection and analysis module is unidirectionally connected to the input ends of the human-computer interaction unit and the model analysis module, the output end of the model analysis module is unidirectionally connected to the input end of the human-computer interaction unit and the slice parameter adjustment unit, and the output end of the human-computer interaction unit is unidirectionally connected to the input end of the slice parameter adjustment unit;

[0016] The data acquisition unit is used to scan the central control chip using laser surface scanning technology and collect the data information obtained by the scanning;

[0017] The data cleaning processing unit is used to clean, denoise and classify the collected data information, and transmit it to the quality inspection and analysis module;

[0018] The quality inspection and analysis module is used to measure the size and surface roughness of the chip, and determine whether the surface has damage and cracks. If there are damage and cracks, the corresponding next step of processing is performed;

[0019] The model analysis module is used to establish and train a machine learning model, analyze the damage on the chip surface through the learning model, determine the cause of the damage, and provide corresponding solutions;

[0020] The slicing parameter adjustment unit is used to control, adjust and detect the parameters of the chip slicer;

[0021] The human-computer interaction unit is used to display the data information of the quality inspection system on the electronic screen in real time through graphics, and the staff can understand the data or adjust the data in real time.

[0022] Furthermore, the quality inspection and analysis module includes a positioning dimension measurement unit, a roughness detection unit, a surface damage detection unit, a crack detection unit, a dimension standard correction unit, a failure calculation unit and a SEM detection unit;

[0023] The input end of the positioning dimension measurement unit is unidirectionally connected to the output end of the data cleaning processing unit, the positive output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the dimension standard correction unit, the negative output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the roughness detection unit, the positive output end of the roughness detection unit is unidirectionally connected to the input end of the failure counting unit, the negative output end of the roughness detection unit is unidirectionally connected to the input end of the surface damage detection unit, the positive output end of the surface damage detection unit is unidirectionally connected to the input end of the model analysis module, the negative output end of the surface damage detection unit is unidirectionally connected to the input end of the crack detection unit, the positive output end of the crack detection unit is unidirectionally connected to the input end of the SEM detection unit, and the failure counting unit and the output end of the SEM detection unit are unidirectionally connected to the input end of the human-computer interaction unit;

[0024] The positioning dimension measurement unit is used to measure and compare the laser scanning data according to the parameter data of the slicing parameter adjustment unit to determine whether the chip size does not meet the requirements. If it does not meet the requirements, it is output to the dimension standard correction unit. If not, that is, it meets the requirements, it is output to the roughness detection unit.

[0025] The roughness detection unit is used to judge whether the roughness of the chip surface does not meet the requirements according to the duration of laser reflection and diffusion, and if not, output it to the failure counting unit, and if it meets, output it to the surface damage detection unit;

[0026] The surface damage detection unit is used to construct an image of the outer surface of the chip based on the data obtained by the laser surface scanning technology, and determine whether the chip surface has damage beyond the specified range. If the detection shows that the chip surface has damage beyond the range, it is transmitted to the model analysis module through the affirmative output terminal; if not, it is output to the crack detection unit;

[0027] The crack detection unit is used to determine whether there is a crack on the chip surface based on the laser reflection and diffusion data, and if there is a crack within a specified range, it is output to the SEM detection unit;

[0028] The size standard correction unit is used to analyze the difference between the current chip size and the standard size, and transmit the correction information to the slice parameter adjustment unit for parameter adjustment;

[0029] The failure calculation unit is used to count the chips with failed surface roughness. When there are multiple chips with failed surface roughness in succession, it means that there may be problems with the processing technology or material quality, and the problem needs to be investigated;

[0030] The SEM detection unit is used to detect cracks on the chip and determine their length, width and depth for subsequent evaluation.

[0031] Furthermore, the model analysis module includes a training set preparation unit, a feature extraction unit, a model analysis unit, and an evaluation and optimization unit;

[0032] The output end of the training set preparation unit is unidirectionally connected to the input end of the feature extraction unit, the input end of the feature extraction unit is unidirectionally connected to the output end of the quality inspection and analysis module, the output end of the feature extraction unit is unidirectionally connected to the input end of the model analysis unit, and the output end of the model analysis unit is unidirectionally connected to the evaluation and optimization unit, the slice parameter adjustment unit and the input end of the human-computer interaction unit;

[0033] The training set preparation unit is used to collect the data sets that have completed analysis to synthesize the data sets;

[0034] The feature extraction unit is used to extract features from the data set to facilitate subsequent model analysis;

[0035] The model analysis unit is used to establish a machine learning model and train it through a training set, so that the model can autonomously analyze and determine whether the damage on the chip surface is scratches, pits, oxidation, burns or wear;

[0036] The evaluation and optimization unit is used to evaluate and optimize the analysis of the model when the model is trained using the training set.

[0037] The beneficial effects of the slicing process of a main control chip of the present invention are as follows:

[0038] By using laser surface scanning technology to collect data, and then using a quality inspection and analysis module to perform multi-faceted automatic quality inspection on the chip's size, surface roughness, surface damage and cracks, the traditional manual small-scale sampling for visual inspection is replaced, solving the existing quality inspection limitations and the problem of difficulty in discovering hidden defects. It also has the advantage of using different coping methods to solve different quality inspection problems, and by using the model analysis module to analyze chip surface damage, it helps workers quickly find the problem that causes the defect. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram of the working process of a slicing quality control system of a slicing process of a master control chip of the present invention;

[0040] Figure 2 It is a schematic diagram of the work flow of the quality inspection and analysis module of the slice quality control system of the present invention;

[0041] Figure 3 The figure is a schematic diagram of the working process of the model analysis module of the slice quality control system of the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] like Figure 1-3 As shown, a slicing process of a main control chip includes the following steps:

[0044] S1: According to the design requirements, silicon wafers that have completed the preliminary process are used as slicing raw materials. The preliminary process includes cleaning, photolithography, exposure and development;

[0045] S2: Coating a protective film on the surface of the silicon wafer covered with photoresist to protect the chip surface from damage or contamination during the cutting process;

[0046] S3: Place the silicon wafer on the slicer table and fix it in the proper position using a clamping device to ensure that the silicon wafer is precisely aligned with the cutting tool to ensure the accuracy and stability of the cutting;

[0047] S4: Use high-precision cutting tools to cut silicon wafers. The movement of the cutting tools is controlled by the slice quality control system to ensure the accuracy of the cutting line and the size of the chip meets the design requirements;

[0048] S5: Cleaning the cut chips to remove the protective film residues and cutting debris that may remain on the surface;

[0049] S6: Grinding and polishing the sliced ​​chips to make their surfaces smoother and more uniform, which helps to improve the quality and reliability of the chips;

[0050] S7: Setting the parameters of the fixed laser, scanning the surface of the chip moving at a uniform speed on the conveyor belt by using the laser surface scanning technology, and collecting data information on the chip surface;

[0051] S8: Analyze the collected data information using the quality inspection analysis module and the model analysis module to determine whether the chip size, surface roughness, surface damage degree and crack degree meet the specified range;

[0052] S9: Qualified products are sent to the next process for packaging, and unqualified products are subjected to different operations according to the reasons for failure;

[0053] The slice quality control system of S4 includes a data acquisition unit, a data cleaning processing unit, a quality inspection and analysis module, a model analysis module, a human-computer interaction unit and a slice parameter adjustment unit.

[0054] Further, the output end of the data acquisition unit is unidirectionally connected to the input end of the data cleaning processing unit, the output end of the data cleaning processing unit is unidirectionally connected to the input end of the quality inspection and analysis module, the output end of the quality inspection and analysis module is unidirectionally connected to the input ends of the human-computer interaction unit and the model analysis module, the output end of the model analysis module is unidirectionally connected to the input end of the human-computer interaction unit and the slice parameter adjustment unit, and the output end of the human-computer interaction unit is unidirectionally connected to the input end of the slice parameter adjustment unit;

[0055] The data acquisition unit is used to scan the central control chip using laser surface scanning technology and collect the data information obtained by the scanning; wherein the laser transmitter is fixed above the conveyor belt, and the chip is conveyed on the conveyor belt after the chip is cut;

[0056] The data cleaning processing unit is used to clean, denoise and classify the collected data information, and transmit it to the quality inspection and analysis module;

[0057] The quality inspection and analysis module is used to measure the size and surface roughness of the chip, and determine whether the surface has damage and cracks. If there are damage and cracks, the corresponding next step of processing is performed;

[0058] The model analysis module is used to establish and train a machine learning model, analyze the damage on the chip surface through the learning model, determine the cause of the damage, and provide corresponding solutions;

[0059] The slicing parameter adjustment unit is used to control, adjust and detect the parameters of the chip slicer;

[0060] The human-computer interaction unit is used to display the data information of the quality inspection system on the electronic screen in real time through graphics, and the staff can understand the data or adjust the data in real time.

[0061] Furthermore, the quality inspection and analysis module includes a positioning dimension measurement unit, a roughness detection unit, a surface damage detection unit, a crack detection unit, a dimension standard correction unit, a failure calculation unit and a SEM detection unit;

[0062] The input end of the positioning dimension measurement unit is unidirectionally connected to the output end of the data cleaning processing unit, the positive output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the dimension standard correction unit, the negative output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the roughness detection unit, the positive output end of the roughness detection unit is unidirectionally connected to the input end of the failure counting unit, the negative output end of the roughness detection unit is unidirectionally connected to the input end of the surface damage detection unit, the positive output end of the surface damage detection unit is unidirectionally connected to the input end of the model analysis module, the negative output end of the surface damage detection unit is unidirectionally connected to the input end of the crack detection unit, the positive output end of the crack detection unit is unidirectionally connected to the input end of the SEM detection unit, and the failure counting unit and the output end of the SEM detection unit are unidirectionally connected to the input end of the human-computer interaction unit;

[0063] The positioning dimension measurement unit is used to measure and compare the laser scanning data according to the parameter data of the slicing parameter adjustment unit, that is, to compare the measured dimension with the set parameter dimension to determine whether the chip dimension does not meet the requirements. If it does not meet the requirements, it is output to the dimension standard correction unit. If not, that is, it meets the requirements, it is output to the roughness detection unit.

[0064] The roughness detection unit is used to calculate the surface roughness according to the duration of laser reflection and diffusion, including common roughness parameters Ra (average roughness), Rz (ten-point average roughness), and Rq (root mean square roughness), to determine whether the roughness of the chip surface meets the requirements. If not, it is output to the failure counting unit, and if it meets, it is output to the surface damage detection unit;

[0065] The surface damage detection unit is used to construct an image of the outer surface of the chip based on the data obtained by the laser surface scanning technology, and determine whether the chip surface has damage beyond the specified range. If the detection shows that the chip surface has damage beyond the range, it is transmitted to the model analysis module through the affirmative output terminal; if not, it is output to the crack detection unit;

[0066] The crack detection unit is used to reconstruct an image based on the data of laser reflection and diffusion, and then identify and mark cracks on the surface by image processing methods such as image enhancement, edge detection, and shape matching, so as to determine whether there are cracks on the chip surface. If there are cracks within a specified range, the cracks are output to the SEM detection unit;

[0067] The size standard correction unit is used to analyze the difference between the current chip size and the standard size, and transmit the correction information to the slice parameter adjustment unit for parameter adjustment;

[0068] The failure calculation unit is used to count the chips that fail the surface roughness test. When there are multiple chips that fail the surface roughness test in succession, there may be problems with the processing technology or material quality, and the problem needs to be investigated. Because when the surface roughness of the chips fails in succession, there may be processing technology problems such as cutting tool wear, too high or too low cutting speed, improper cutting fluid, or raw material quality problems such as poor surface quality of the raw materials, pores or inclusions inside the materials, or tool wear problems such as the use of tools that have been severely worn, or processing environment problems such as too high humidity in the processing environment, which leads to oxidation or corrosion on the processing surface.

[0069] The SEM detection unit is used to detect cracks on the chip using a scanning electron microscope and determine their length, width and depth, so as to facilitate subsequent evaluation by staff; specifically, by adjusting the parameters of the SEM (such as electron beam energy, scanning speed, detector settings, etc.), a high-resolution, high-contrast sample image can be obtained. After obtaining the surface image of the sample, the SEM is usually equipped with image processing and analysis software, which can be used to observe, measure and analyze the cracks on the sample. The cracks are usually clearly visible in the image and can be preliminarily evaluated based on their shape, size and depth.

[0070] Furthermore, the model analysis module includes a training set preparation unit, a feature extraction unit, a model analysis unit, and an evaluation and optimization unit;

[0071] The output end of the training set preparation unit is unidirectionally connected to the input end of the feature extraction unit, the input end of the feature extraction unit is unidirectionally connected to the output end of the quality inspection and analysis module, the output end of the feature extraction unit is unidirectionally connected to the input end of the model analysis unit, and the output end of the model analysis unit is unidirectionally connected to the evaluation and optimization unit, the slice parameter adjustment unit and the input end of the human-computer interaction unit;

[0072] The training set preparation unit is used to collect the data sets that have completed analysis to synthesize the data sets;

[0073] The feature extraction unit is used to extract features from the data set to facilitate subsequent model analysis;

[0074] The model analysis unit is used to establish a machine learning model and train it through a training set, so that the model can autonomously analyze and determine whether the damage on the chip surface is scratches, pits, oxidation, burns or wear, and determine which factor among processing technology problems, material quality problems, tool wear and environmental factors may cause the damage;

[0075] The evaluation and optimization unit is used to evaluate and optimize the analysis of the model when the model is trained using the training set.

[0076] Embodiment 1:

[0077] S1: According to the design requirements, silicon wafers that have completed the preliminary process are used as slicing raw materials. The preliminary process includes cleaning, photolithography, exposure and development;

[0078] S2: Coating a protective film on the surface of the silicon wafer covered with photoresist to protect the chip surface from damage or contamination during the cutting process;

[0079] S3: Place the silicon wafer on the slicer table and fix it in the proper position using a clamping device to ensure that the silicon wafer is precisely aligned with the cutting tool to ensure the accuracy and stability of the cutting;

[0080] S4: Use high-precision cutting tools to cut silicon wafers. The movement of the cutting tools is controlled by the slice quality control system to ensure the accuracy of the cutting line and the size of the chip meets the design requirements;

[0081] S5: Cleaning the cut chips to remove the protective film residues and cutting debris that may remain on the surface;

[0082] S6: Grinding and polishing the sliced ​​chips to make their surfaces smoother and more uniform, which helps to improve the quality and reliability of the chips;

[0083] S7: Setting the parameters of the fixed laser, scanning the surface of the chip moving at a uniform speed on the conveyor belt by using the laser surface scanning technology, and collecting data information on the chip surface;

[0084] S8: Analyze the collected data information using the quality inspection analysis module and the model analysis module to determine whether the chip size, surface roughness, surface damage degree and crack degree meet the specified range;

[0085] S9: Qualified products are sent to the next process for packaging, and unqualified products are subjected to different operations according to the reasons for failure;

[0086] The present invention uses a series of steps and equipment to carry out the slicing process of the main control chip, including the preparation of silicon wafers, lamination, cutting, grinding and polishing, so that the slicing process of the main control chip removes the quality inspection restrictions and makes the quality control process more perfect.

[0087] The specific implementation of the present invention is as follows:

[0088] When the quality control process is to be carried out after the pre-process of the slicing process is completed, the parameters of the laser generator are set, and the conveyor belt passes the chips on it through the laser generator at a uniform speed. The data information on the laser receiver is collected by the data acquisition unit. The data cleaning processing unit transmits the data to the quality inspection analysis module after processing. The positioning size measurement unit measures and compares the laser scanning data according to the parameter data of the slicing parameter adjustment unit to determine whether the chip size does not meet the requirements. If it does not meet the requirements, it is output to the size standard correction unit. The size standard correction unit analyzes the difference between the current chip size and the standard size, and transmits the correction information to the slicing parameter adjustment unit for The parameters are adjusted. If not, that is, it meets the requirements, it is output to the roughness detection unit; the roughness detection unit determines whether the roughness of the chip surface does not meet the requirements according to the duration of laser reflection and diffusion. If it does not meet the requirements, it is output to the unqualified counting unit. The unqualified calculation unit counts the chips with unqualified surface roughness. When there are multiple chips with unqualified surface roughness in succession, it may be a problem of processing technology or material quality. The human-computer interaction unit allows the staff to troubleshoot the problem. If it meets the requirements, it is output to the surface damage detection unit; the surface damage detection unit performs image grouping on the outer surface of the chip according to the data obtained by the laser surface scanning technology. The chip surface is constructed and judged whether there is damage beyond the specified range. If the detection shows that there is damage beyond the range on the chip surface, it is transmitted to the model analysis module through the affirmative output end. The model analysis module analyzes the damage on the chip surface through the learning model, judges what causes the damage, and provides corresponding solutions. If there is no damage, it is output to the crack detection unit; the crack detection unit judges whether there is a crack on the chip surface according to the data of laser reflection and diffusion. If there is a crack within the specified range, it is output to the SEM detection unit. The SEM detection unit detects the crack on the chip and determines its length, width and depth, which is convenient for subsequent evaluation. ; When any of the positioning dimension measurement unit, the roughness detection unit, the surface damage detection unit, and the crack detection unit of the quality inspection and analysis module detects parameters that do not meet the requirements, the chip quality inspection fails and the corresponding correction steps are performed; when damage occurs on the chip surface, the data information is transmitted to the model analysis module, and the model analysis unit in the model analysis module autonomously analyzes the damage on the chip surface through the established machine learning model to determine whether it is scratches, pits, oxidation, burns or wear, and proposes corresponding processing or adjustment methods to be transmitted to the human-computer interaction unit; when the chip passes the inspection of all inspection units, it means that the chip has passed the quality inspection.By setting up laser surface scanning technology to collect data, the quality inspection and analysis module is used to perform automatic quality inspection on the chip size, surface roughness, surface damage and cracks in various aspects, and if the quality inspection fails, there will be corresponding solutions or adjustments. This solves the problem mentioned in the background technology that "the current quality inspection process after the chip slicing process is usually a sampling inspection of a small part of a large batch of products, which has quality inspection limitations, and most of the obvious defects are found through manual visual inspection, but some hidden defects may be difficult to detect."

[0089] The present invention and its embodiments are described above, which is not restrictive. What is shown in the accompanying drawings is only one of the embodiments of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in this field are inspired by it and do not deviate from the purpose of the invention, they can creatively design structural methods and embodiments similar to the technical scheme, which should all fall within the scope of protection of the present invention.

Claims

1. A slicing process for a master control chip, characterized in that: The following steps are involved: S1: According to the design requirements, silicon wafers that have completed the preliminary process are used as slicing raw materials. The preliminary process includes cleaning, photolithography, exposure and development; S2: Coating a protective film on the surface of the silicon wafer covered with photoresist to protect the chip surface from damage or contamination during the cutting process; S3: Place the silicon wafer on the slicer table and fix it in the proper position using a clamping device to ensure that the silicon wafer is precisely aligned with the cutting tool to ensure the accuracy and stability of the cutting; S4: Use a high-precision cutting tool to cut the silicon wafer. The movement of the cutting tool is controlled by a slicing quality control system to ensure that the accuracy of the cutting line and the size of the chip meet the design requirements. The slicing quality control system includes a data acquisition unit, a data cleaning processing unit, a quality inspection and analysis module, a model analysis module, a human-computer interaction unit and a slicing parameter adjustment unit. The quality inspection and analysis module includes a positioning dimension measurement unit, a roughness detection unit, a surface damage detection unit, a crack detection unit, a dimension standard correction unit, a failure calculation unit and a SEM detection unit. The model analysis module includes a training set preparation unit, a feature extraction unit, a model analysis unit and an evaluation optimization unit; S5: Cleaning the cut chips to remove the protective film residues and cutting debris that may remain on the surface; S6: Grinding and polishing the sliced ​​chips to make their surfaces smoother and more uniform, which helps to improve the quality and reliability of the chips; S7: Setting the parameters of the fixed laser, scanning the surface of the chip moving at a uniform speed on the conveyor belt by using the laser surface scanning technology, and collecting data information on the chip surface; S8: Analyze the collected data information using the quality inspection analysis module and the model analysis module to determine whether the chip size, surface roughness, surface damage degree and crack degree meet the specified range; S9: Qualified products are sent to the next process for packaging, and unqualified products are subjected to different operations based on the reasons for failure.

2. The slicing process of a main control chip according to claim 1, characterized in that: The output end of the data acquisition unit is unidirectionally connected to the input end of the data cleaning processing unit, the output end of the data cleaning processing unit is unidirectionally connected to the input end of the quality inspection and analysis module, the output end of the quality inspection and analysis module is unidirectionally connected to the input ends of the human-computer interaction unit and the model analysis module, the output end of the model analysis module is unidirectionally connected to the input end of the human-computer interaction unit and the slice parameter adjustment unit, and the output end of the human-computer interaction unit is unidirectionally connected to the input end of the slice parameter adjustment unit; The data acquisition unit is used to scan the central control chip using laser surface scanning technology and collect the data information obtained by the scanning; The data cleaning processing unit is used to clean, denoise and classify the collected data information, and transmit it to the quality inspection and analysis module; The quality inspection and analysis module is used to measure the size and surface roughness of the chip, and determine whether the surface is damaged or cracked. If there is damage or crack, the corresponding next step is performed; The model analysis module is used to establish and train a machine learning model, analyze the damage on the chip surface through the learning model, determine the cause of the damage, and provide corresponding solutions; The slicing parameter adjustment unit is used to control, adjust and detect the parameters of the chip slicer; The human-computer interaction unit is used to display the data information of the quality inspection system on the electronic screen in real time through graphics, and the staff can understand the data or adjust the data in real time.

3. The slicing process of a main control chip according to claim 2, characterized in that: The input end of the positioning dimension measurement unit is unidirectionally connected to the output end of the data cleaning processing unit, the positive output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the dimension standard correction unit, the output end of the dimension standard correction unit is unidirectionally connected to the input end of the slicing parameter adjustment unit, the negative output end of the positioning dimension measurement unit is unidirectionally connected to the input end of the roughness detection unit, the positive output end of the roughness detection unit is unidirectionally connected to the input end of the failure calculation unit, the negative output end of the roughness detection unit is unidirectionally connected to the input end of the surface damage detection unit, the positive output end of the surface damage detection unit is unidirectionally connected to the input end of the model analysis module, the negative output end of the surface damage detection unit is unidirectionally connected to the input end of the crack detection unit, the positive output end of the crack detection unit is unidirectionally connected to the input end of the SEM detection unit, the output end of the failure calculation unit, the output end of the SEM detection unit and the output end of the model analysis module are unidirectionally connected to the input end of the human-computer interaction unit respectively; The positioning dimension measurement unit is used to measure and compare the laser scanning data according to the parameter data of the slicing parameter adjustment unit to determine whether the chip size does not meet the requirements. If it does not meet the requirements, it is output to the dimension standard correction unit. If not, that is, it meets the requirements, it is output to the roughness detection unit. The roughness detection unit is used to judge whether the roughness of the chip surface does not meet the requirements according to the duration of laser reflection and diffusion, and if not, output it to the failure calculation unit, and if it meets, output it to the surface damage detection unit; The surface damage detection unit is used to construct an image of the outer surface of the chip based on the data obtained by the laser surface scanning technology, and determine whether the chip surface has damage beyond the specified range. If the detection shows that the chip surface has damage beyond the range, it is transmitted to the model analysis module through the affirmative output terminal for further analysis. If it does not exist, it is output to the crack detection unit; The crack detection unit is used to determine whether there is a crack on the chip surface based on the laser reflection and diffusion data, and if there is a crack within a specified range, it is output to the SEM detection unit; The size standard correction unit is used to analyze the difference between the current chip size and the standard size, and transmit the correction information to the slice parameter adjustment unit for parameter adjustment; The failure calculation unit is used to count the chips with failed surface roughness. When there are multiple chips with failed surface roughness in succession, it means that there may be problems with the processing technology or material quality, and the problem needs to be investigated; The SEM detection unit is used to detect cracks on the chip and determine their length, width and depth for subsequent evaluation.

4. The slicing process of a main control chip according to claim 2, characterized in that: The output end of the training set preparation unit is unidirectionally connected to the input end of the feature extraction unit, the input end of the feature extraction unit is unidirectionally connected to the output end of the quality inspection and analysis module, the output end of the feature extraction unit is unidirectionally connected to the input end of the model analysis unit, and the output end of the model analysis unit is unidirectionally connected to the evaluation and optimization unit, the slice parameter adjustment unit and the input end of the human-computer interaction unit; The training set preparation unit is used to collect the data sets that have completed analysis to synthesize the data sets; The feature extraction unit is used to extract features from the data set to facilitate subsequent model analysis; The model analysis unit is used to establish a machine learning model and train it through a training set, so that the model can autonomously analyze and determine whether the damage on the chip surface is scratches, pits, oxidation, burns or wear; The evaluation and optimization unit is used to evaluate and optimize the analysis of the model when the model is trained using the training set.

Images