Evaluation method and system for multi-layer chip stacking scheme

Abstract

The invention relates to the technical field of stack packaging process optimization, in particular to an evaluation method and system for a multi-layer chip stacking scheme, and the method comprises the steps: carrying out the data collection of each determined scheme variable item, obtaining a plurality of variable information sets, selecting the process information for each variable scheme, obtaining a plurality of basic process schemes, by directionally adjusting the basic process scheme, acquiring the pre-selected lamination scheme, performing process digital simulation on the respective lamination schemes, and acquiring the optimal lamination scheme according to the simulation result, the problem of low process debugging efficiency caused by multivariable coupling in the lamination packaging process is effectively solved, and the process debugging efficiency is improved. By performing data acquisition and compatibility evaluation on a plurality of scheme variable items, the scheme can comprehensively analyze the mutual influence of each process parameter in different lamination schemes, so that the most suitable lamination scheme is more accurately selected, and the production efficiency and the product quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of optimizing the stacked packaging process, and particularly to an evaluation method and system for a multi-layer chip stacking scheme. Background Art

[0002] With the rapid development of semiconductor packaging technology towards high density and high integration, the stacked packaging process, as a key technology for realizing three-dimensional integration and miniaturized packaging, has become one of the core research directions in the field of advanced packaging. Through multi-chip stacking and interconnection, the stacked packaging technology significantly improves device performance and reduces the packaging volume, and is widely used in mobile devices, high-performance computing, Internet of Things and other fields. Especially driven by emerging demands such as 5G communication and artificial intelligence chips, the stacked packaging process puts forward higher requirements for the precision, reliability and efficiency of the stacking scheme, prompting the industry to explore intelligent process optimization methods that combine data-driven and virtual simulation.

[0003] When selecting a stacking scheme in the existing stacked packaging process, it often relies on experience or a single parameter optimization method, and cannot comprehensively consider the influence of all variable factors. Traditional stacking scheme selection methods usually based on historical data and empirical formulas, but these methods lack comprehensive digital simulation and precise tuning of the process, and are difficult to cope with complex packaging requirements and changing production environments. At the same time, there is also a lack of systematic evaluation of the compatibility and mutual influence between various process parameters. In practical applications, the selection of scheme variables and process adjustment often have uncertainties, which not only reduces process efficiency, but also may lead to unstable product quality.

[0004] The information disclosed in this background art section is only intended to deepen the understanding of the overall background art of the present disclosure, and should not be regarded as an admission or any form of suggestion that this information constitutes the prior art known to those skilled in the art. Summary of the Invention

[0005] The present invention provides an evaluation method and system for a multi-layer chip stacking scheme, which can effectively solve the problems in the background art.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: An evaluation method for a multi-layer chip stacking scheme, the method comprising: Determine the scheme variable items, and collect data for each of the scheme variable items to obtain a number of variable information sets, and the scheme variable items constitute the stacking scheme in the stacked packaging process; Select the process information corresponding to each of the scheme variable items from the variable information set corresponding thereto to obtain a number of basic process schemes; Directionally adjust the process information of the program variable items in the basic process plan to obtain several optimized process plans, and use several of the optimized process plans and the basic process plan as preselected lamination plans; Construct a virtual process mirror, and perform process digital simulation on the preselected lamination plans respectively to obtain lamination plan feedback; Select the corresponding preselected lamination plan as the best lamination plan according to the lamination plan feedback.

[0007] Furthermore, each program variable item selects the process information corresponding to the variable information set and obtains several basic process plans, including: Establish a lamination process database, extract historical lamination plans according to the lamination process database, and mark the program variable items corresponding to the historical lamination plans as preselected combination plans. The program variable items include glue type information, dispensing path information, and lamination geometry information; Extract several dispensing path information and lamination geometry information according to the lamination process database, and evaluate the compatibility of each dispensing path information with several lamination geometry information respectively; Classify the historical lamination plans according to the glue type information to obtain several historical glue lamination plan sets, where each historical glue lamination plan set corresponds to each glue type information one by one; Set the call weight according to the compatibility evaluation result and the preselected combination plan, and recombine the process information of the remaining program variable items of each historical glue lamination plan set according to the call weight to obtain several basic process plans.

[0008] Furthermore, evaluate the compatibility of several dispensing path information with several lamination geometry information respectively, including: Extract the interlayer geometry parameters according to the lamination geometry information, and determine the geometric process threshold based on the packaging process requirements; Calculate the geometric adaptation range of the dispensing path information, obtain the glue amount distribution constraint according to the packaging process requirements, and screen the lamination geometry information based on the glue amount distribution constraint; Detect whether the edge distance between the projection area of the dispensing path information in the lamination packaging direction and the lamination geometry information meets the preset safety threshold, and obtain the path conflict detection result according to the distance detection result; Generate the compatibility evaluation result of the dispensing path information and the lamination geometry information according to the geometric adaptation range, geometric process threshold, and path conflict detection result.

[0009] Furthermore, directionally adjust the process information of the program variable items in the basic process plan to obtain several optimized process plans, including: S1: Extract a number of recombinant glue stack schemes sets according to the basic process scheme, perform information migration on the recombinant glue stack schemes sets based on the compatibility evaluation results, mark and eliminate the schemes that are the same as the basic process scheme after label exchange, and obtain the exchanged process scheme. The information migration is the exchange of the corresponding process information of the scheme variable items between different recombinant glue stack schemes sets; S2: Directionally adjust a single scheme variable item in the exchanged process scheme, and re-evaluate the compatibility of the exchanged process scheme after directional adjustment to obtain an optimized process scheme; Use the optimized process scheme as the optimized basic process scheme and repeat steps S1 and S2 until the preset number of iterations is reached to obtain a number of optimized process schemes.

[0010] Further, set the call weights according to the compatibility evaluation results and the preselected combination schemes, including: Generate a compatibility score for each preselected combination scheme based on the compatibility evaluation results; Extract the process success rate of the historical lamination scheme according to the lamination process database, and assign a basic weight to the historical scheme variable items according to the process success rate; Perform linear weighting on the compatibility score and the basic weight to obtain the call weight.

[0011] Further, perform process digital simulation on the preselected lamination schemes respectively to obtain lamination scheme feedback, including: Extract the encapsulation material parameters according to the preselected lamination scheme, and construct a virtual process mirror. The virtual process mirror is divided into a geometric structure layer and a process parameter layer; Obtain the encapsulation structure parameters according to the geometric structure layer, and obtain the glue type and curing conditions according to the process parameter layer; Perform glue flow simulation according to the encapsulation structure parameters, glue type and curing conditions to obtain the glue distribution characteristics, and calculate the thermo-mechanical properties based on the process parameter layer; Evaluate the preselected lamination scheme according to the glue distribution characteristics and thermo-mechanical properties to obtain lamination scheme feedback.

[0012] Further, select the corresponding preselected lamination scheme as the best lamination scheme according to the lamination scheme feedback, including: Generate a call score according to the glue distribution characteristics and thermo-mechanical properties; Set a performance evaluation threshold according to the historical lamination scheme, compare the call score with the performance evaluation threshold. If the call score exceeds the performance evaluation threshold, mark the corresponding preselected lamination scheme as a lamination candidate scheme; The candidate stacking solutions are sorted to obtain the best stacking solution.

[0013] Furthermore, the candidate stacking solutions are sorted to obtain the best stacking solution, including: S3: marking any of the stacking candidate solutions as a sequence base point, comparing the remaining stacking candidate solutions with the sequence base point respectively, adjusting the relative positions of the remaining stacking candidate solutions and the sequence base point according to the comparison results, and obtaining two stacking unordered sets; S4: canceling the mark of the stacking candidate solution, and selecting any stacking candidate solution in the stacking unordered set as the sequence base point, and comparing it with the remaining stacking candidate solutions in the stacking unordered set, and adjusting the relative positions respectively; Repeat steps S3 and S4 until the stacking candidate solutions are sorted and the best stacking solution is obtained.

[0014] An evaluation system for a multi-layer chip stacking solution, the system comprising: A variable collection and integration module determines the scheme variable items and collects data for each of the scheme variable items to obtain a plurality of variable information sets, wherein the scheme variable items constitute a stacking scheme in a stacking packaging process; A process scheme generating module, wherein each of the scheme variable items selects the process information corresponding to the variable information set and obtains a number of basic process schemes; A process scheme tuning module is used to adjust the process information of the scheme variable items in the basic process scheme in a targeted manner, obtain a number of optimized process schemes, and use the several optimized process schemes and the basic process scheme as pre-selected lamination schemes; A virtual simulation evaluation module is used to construct a virtual process image, perform digital process simulation on the pre-selected stacking schemes, and obtain stacking scheme feedback; The scheme optimization decision module selects the corresponding pre-selected stacking scheme as the best stacking scheme according to the stacking scheme feedback.

[0015] Furthermore, the virtual simulation evaluation module includes: A virtual image building unit, which extracts packaging material parameters according to the pre-selected lamination scheme and builds a virtual process image, wherein the virtual process image is divided into a geometric structure layer and a process parameter layer; A parameter extraction and analysis unit, which obtains packaging structure parameters according to the geometric structure layer, and obtains glue type and curing conditions according to the process parameter layer; A fluid-heat coupling simulation unit performs glue flow simulation according to the packaging structure parameters, glue type and curing conditions, obtains glue distribution characteristics, and calculates thermomechanical properties based on the process parameter layer; The solution performance evaluation unit evaluates the preselected lamination solution according to the glue distribution characteristics and thermomechanical properties, and obtains the lamination solution feedback.

[0016] Through the technical solution of the present invention, the following technical effects can be achieved: Effectively solve the problems of low debugging efficiency, insufficient utilization of historical data, and poor process stability caused by multi-variable coupling in the stacking packaging process. By means of data acquisition, virtual process simulation, and iterative optimization, dynamically adjust the process parameters of the lamination solution, greatly improving the accuracy and adaptability of the process solution, enabling the finally selected lamination solution to more accurately meet different packaging requirements and production environments. This optimization process significantly improves the stability of the packaging process and product quality, reduces the defective rate in the production process, and at the same time helps enterprises reduce production costs and improve production efficiency.

[0017] The above description is only an overview of the technical solution of this application. In order to be able to more clearly understand the technical means of this application, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0019] Figure 1 It is a flow schematic diagram of the evaluation method for the multi-layer chip stacking solution; Figure 2 It is a flow schematic diagram of obtaining the basic process solution; Figure 3 It is a flow schematic diagram of compatibility evaluation; Figure 4 It is a flow schematic diagram of obtaining the optimized process solution; Figure 5 It is a structural schematic diagram of the evaluation system for the multi-layer chip stacking solution. Detailed Description of the Invention

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the description of this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0022] Embodiment 1; As Figure 1 shown, this application provides an evaluation method for a multi-layer chip stacking scheme, and the method includes: S100: Determine the scheme variable items, collect data for each scheme variable item, and obtain several variable information sets. The scheme variable items form the stacking scheme in the stacked die packaging process; S200: Select the process information in the corresponding variable information set for each scheme variable item and obtain several basic process schemes; S300: Directionally adjust the process information of the scheme variable items in the basic process schemes to obtain several optimized process schemes, and use the several optimized process schemes and the basic process schemes as preliminary stacking schemes; S400: Construct a virtual process mirror, perform process digital simulation on the preliminary stacking schemes respectively, and obtain stacking scheme feedback; S500: Select the corresponding preliminary stacking scheme as the best stacking scheme according to the stacking scheme feedback.

[0023] Specifically, first, by analyzing the key factors affecting the performance of the stacked package process, such as the characteristics of the bonding material, the design of the coating path, etc., combined with historical process data and actual production requirements, the scheme variable items are determined. Subsequently, through automated equipment, sensors or manual operations, the relevant data of each scheme variable item are collected and stored in a specific format to form several variable information sets. Then, each scheme variable item selects the corresponding process information according to its data set and generates several basic process schemes. For example, according to the characteristics of the bonding material, a bonding material with high heat resistance or fast curing characteristics can be selected. Based on the coating path information, the appropriate coating equipment and coating method are selected. On this basis, the scheme variable items are adjusted directionally to generate multiple optimized process schemes. By repeatedly adjusting the process parameters, it is ensured that the finally selected stacking scheme can adapt to different packaging requirements. Then, a virtual process mirror is constructed to digitally simulate the process of the selected preliminary stacking scheme. Through simulation, problems that may occur during the packaging process, such as the uniform distribution of glue, interlayer contact, and thermal stress, can be detected in advance, thus effectively reducing the risk of process failure in production. Finally, according to the results feedback by the virtual process mirror, the performance indicators of the stacking scheme, such as the glue distribution characteristics and thermo-mechanical properties, are comprehensively evaluated to select the best stacking scheme.

[0024] Through the technical solution of the present invention, the problems of low debugging efficiency, insufficient utilization of historical data, and poor process stability caused by multi-variable coupling in the stacked package process are effectively solved. By means of data collection, virtual process simulation, and iterative optimization, the process parameters of the stacking scheme are dynamically adjusted, greatly improving the accuracy and adaptability of the process scheme, so that the finally selected stacking scheme can more accurately meet different packaging requirements and production environments. This optimization process significantly improves the stability of the packaging process and product quality, reduces the defective rate in the production process, and at the same time helps enterprises reduce production costs and improve production efficiency.

[0025] Furthermore, as Figure 2 shown, each scheme variable item selects the process information in the corresponding variable information set and obtains several basic process schemes, including: S210: Establish a stacking process database, extract historical stacking schemes from the stacking process database, and mark the scheme variable items corresponding to the historical stacking schemes as preliminary combination schemes. The scheme variable items include glue type information, dispensing path information, and stacking geometry information; S220: Extract several dispensing path information and stacking geometry information from the stacking process database, and evaluate the compatibility of each dispensing path information with several stacking geometry information respectively; S230: Classify the historical lamination schemes according to the glue type information to obtain several historical glue lamination scheme sets, where each historical glue lamination scheme set corresponds one-to-one to each glue type information; S240: Set call weights according to the compatibility evaluation results and the preselected combination schemes, and reorganize the process information of the remaining scheme variable items of each historical glue lamination scheme set according to the call weights to obtain several basic process schemes.

[0026] As an optimization of the above embodiment, first collect and organize various lamination schemes and their related process parameters (such as glue type, dispensing path, lamination geometry information, etc.) used in the historical production process, classify and label the process information of each scheme, and construct a lamination process database. This database not only records the correspondence between different glue types and lamination geometry configurations, but also includes data such as the actual performance and success rate of each process scheme. According to the lamination process database, each historical lamination scheme will be marked with its corresponding scheme variable items to form a preselected combination scheme. For example, a historical lamination scheme may include using a certain type of glue, a specific dispensing path, and a specific lamination geometry configuration. Then, extract several dispensing path information and lamination geometry information from the lamination process database. The dispensing path information includes the distribution path of the glue during the lamination process, and the lamination geometry information includes the shape, size, arrangement method, etc. of the lamination layers. Subsequently, perform a compatibility evaluation on each dispensing path information and multiple lamination geometry information to determine whether they can be mutually adapted. For example, a specific dispensing path may require a large glue coverage area, while some lamination geometry designs may result in uneven glue coverage, so the two may not be compatible. At this time, through the compatibility evaluation, select the best-matched dispensing path and lamination geometry configuration. To further optimize the process scheme, it is necessary to classify the historical lamination schemes according to the glue type. Each glue type corresponds to a set of historical glue lamination scheme sets. These historical glue lamination scheme sets contain all the lamination schemes using this glue type, and each historical scheme set contains the dispensing path and lamination geometry information related to this glue type. After obtaining the basic data through the compatibility evaluation results and the preselected combination schemes, next, set call weights for each preselected combination scheme according to the results of the compatibility evaluation. This weight is used to evaluate the adaptability of each historical lamination scheme to ensure that the finally selected scheme has a high process success rate and compatibility. Through the call weight, the process information of the remaining scheme variable items in each historical glue lamination scheme set can be reorganized. Specifically, the scheme with a higher call weight will receive more attention in the next process optimization to ensure that the optimized process scheme has better performance. Through the above steps, several basic process schemes can finally be obtained.

[0027] Furthermore, as Figure 3As shown, the compatibility of a number of dispensing path information with a number of stacked chip geometry information is evaluated, including: S221: Extract the interlayer geometry parameters according to the stacked chip geometry information, and determine the geometric process threshold based on the packaging process requirements; S222: Calculate the geometric adaptation range of the dispensing path information, obtain the glue volume distribution constraint according to the packaging process requirements, and screen the stacked chip geometry information based on the glue volume distribution constraint; S223: Detect whether the edge spacing between the projection area of the dispensing path information in the stacking packaging direction and the stacked chip geometry information meets the preset safety threshold, and obtain the path conflict detection result according to the spacing detection result; S224: Generate the compatibility evaluation result of the dispensing path information and the stacked chip geometry information according to the geometric adaptation range, geometric process threshold and path conflict detection result.

[0028] In this embodiment, first, the geometric parameters of each layer are extracted from the stacked chip geometry information. These parameters include the thickness, size, shape and relative position of each layer. Through these geometric parameters, the spacing between each layer and the interlayer geometric relationship can be calculated. In some embodiments, the calculation method is as follows: First, calculate the bottom and top positions of adjacent layers to obtain the vertical distance between layers (i.e., the interlayer spacing). In addition, it is also necessary to calculate the geometric relationship between layers, such as whether the layers are aligned, whether there is an offset and angular deviation. The geometric center offset of the layer can be determined by calculating the difference between the geometric center coordinates of the two layers, and the angular relationship is represented by calculating the relative rotation angle. The interlayer geometric relationship is obtained according to the calculation results. At the same time, in order to ensure the effectiveness of the stacked chip process, it is necessary to determine the geometric process threshold based on the specific requirements of the packaging process; for example, in some high-precision packaging processes, in order to avoid glue overflow or uneven distribution caused by too small interlayer spacing, a minimum interlayer spacing threshold may need to be set. These thresholds are used to limit the range of geometric parameters to ensure that each layer can be correctly stacked during the stacked chip process; subsequently, for each dispensing path, calculate its geometric adaptation range: A simulation tool based on fluid dynamics can be used to calculate the geometric adaptation range to ensure the optimal cooperation between the dispensing path and the geometric structure; then, in order to ensure the compatibility between the dispensing path and the stacked chip geometry information, it is also necessary to detect the projection area of the dispensing path in the stacking packaging direction. If the projection area of the dispensing path is too close to the edge of the stacked chip geometry information, it may cause glue overflow or uneven coating, thus affecting the quality of the packaging; after completing steps such as geometric adaptation range calculation, glue volume distribution constraint screening, and path conflict detection, according to all evaluation results, generate the compatibility evaluation result of the dispensing path information and the stacked chip geometry information.

[0029] Furthermore, as Figure 4As shown in the figure, the process information of the scheme variable items in the basic process scheme for directional adjustment is obtained, and several optimized process schemes are obtained, including: S1: Extract several sets of recombined glue stack schemes from the basic process scheme, perform information migration on the sets of recombined glue stack schemes based on the compatibility evaluation results, mark and screen out the schemes that are the same as the basic process scheme after the exchange, and obtain the exchanged process schemes. The information migration is the exchange of the corresponding process information of the scheme variable items between different sets of recombined glue stack schemes; S2: Directionally adjust a single scheme variable item in the exchanged process scheme, and re-evaluate the compatibility of the exchanged process scheme after the directional adjustment to obtain the optimized process scheme; Take the optimized process scheme as the optimized basic process scheme and repeat steps S1 and S2 until the preset number of iterations is reached to obtain several optimized process schemes.

[0030] Specifically, first, several sets of recombined glue stack schemes are extracted from the basic process scheme. Each set includes multiple process variable items related to glue type, dispensing path, lamination geometry, etc. The sets of recombined glue stack schemes are subjected to information migration based on the compatibility evaluation results. The information migration refers to the exchange of the corresponding process information between different sets of recombined glue stack schemes. For example, there are two different glue types, and each glue type corresponds to several lamination schemes. According to the compatibility evaluation results, the dispensing paths or lamination geometry information of the lamination schemes of different glue types are migrated. And there may be schemes that are the same as the basic process scheme, and these schemes are screened out to obtain several exchanged process schemes. After obtaining the exchanged process schemes, next, a single scheme variable item in the exchanged process scheme is directionally adjusted. The purpose of this step is to refine and adjust the key variable items that may affect the encapsulation effect in the preliminarily recombined scheme. For example, if it is found that the glue flow is uneven or the curing speed is too fast in the matching of the dispensing path and geometric design, the shape of the dispensing path can be adjusted, or the flow rate of the glue can be adjusted to ensure the stability and accuracy of the encapsulation process. However, after each adjustment, the compatibility between each scheme variable item needs to be re-evaluated to ensure that the adjusted scheme can meet the overall process requirements. After the operations of steps S1 and S2, the number of iterations can be set according to historical experience, and steps S1 and S2 are repeatedly executed until the process scheme reaches the preset optimization standard or the upper limit of the number of iterations. In each iteration process, the optimized process scheme will be used as the new basic process scheme and input into the next round of optimization to continuously improve the accuracy and adaptability of the process.

[0031] Furthermore, call weights are set according to the compatibility evaluation results and the preselected combination schemes, including: Based on the compatibility evaluation results, generate a compatibility score for each preselected combination scheme; Extract the process success rate of historical lamination schemes from the lamination process database, and assign a basic weight to the historical scheme variable items according to the process success rate; Perform linear weighting on the compatibility score and the basic weight to obtain the call weight.

[0032] As an optimization of the above embodiment, according to the compatibility evaluation results of each scheme variable item, each scheme generates a compatibility score based on the matching degree of these parameters, which reflects the applicability and potential process efficiency of the scheme. The generation of the compatibility score can utilize advanced analysis algorithms, such as machine learning-based prediction models; Subsequently, extract the process success rate of historical lamination schemes from the lamination process database: The extraction of historical success rates can be combined with big data analysis methods. In some embodiments, statistical analysis methods are used to calculate the long-term and short-term success rates of the scheme. For the evaluation of historical processes, the weighted average method can be used to comprehensively consider the influence of each historical data point, and the success rate can be dynamically adjusted according to the change trend of process parameters and different production conditions. By weighting the success rates under different production environments, raw materials, and equipment conditions, the adaptability and reliability of each scheme can be more accurately reflected. This data shows the performance and reliability of each scheme in actual applications; Each historical scheme variable item is assigned a basic weight according to its success rate, which reflects the actual effect and credibility of the scheme. The determination of the basic weight should consider the usage frequency of the scheme and its performance under different conditions, and statistical analysis methods are used to comprehensively evaluate the long-term and short-term success rates of the scheme, as well as its adaptability under different production conditions; Finally, perform linear weighting on the compatibility score and the basic weight to generate the final call weight, which is the key factor in determining whether to adopt a certain scheme. A scheme with a high weight indicates good compatibility and a high historical success rate.

[0033] Furthermore, perform process digital simulation on the preselected lamination schemes respectively to obtain lamination scheme feedback, including: Extract the encapsulation material parameters according to the preselected lamination scheme, and construct a virtual process mirror, which is divided into a geometric structure layer and a process parameter layer; Obtain the encapsulation structure parameters according to the geometric structure layer, and obtain the glue type and curing conditions according to the process parameter layer; Perform glue flow simulation according to the encapsulation structure parameters, glue type, and curing conditions to obtain the glue distribution characteristics, and calculate the thermo-mechanical properties based on the process parameter layer; Evaluate the preselected lamination scheme according to the glue distribution characteristics and thermo-mechanical properties to obtain lamination scheme feedback.

[0034] In this embodiment, according to the preselected lamination scheme, relevant parameters of the encapsulation material are extracted, mainly including glue type, curing conditions, and geometric information of the lamination. These parameters are the basis for constructing the virtual process mirror. The virtual process mirror consists of two parts: a geometric structure layer and a process parameter layer. A 3D modeling tool can be used to generate the geometric structure layer, and the process parameter layer can be defined through hydrodynamic simulation and thermodynamic models. The geometric structure layer is used to describe the geometric shape, size, interlayer spacing, arrangement method, etc. of the encapsulation structure. The process parameter layer contains process parameters such as glue type, curing temperature, and curing time. These parameters directly affect the fluidity and curing process of the glue, and thus affect the overall performance of the encapsulation. Then, detailed parameters of the encapsulation structure are extracted through the geometric structure layer, such as the thickness, size, alignment method, interlayer spacing, etc. of each layer. At the same time, the glue type (such as epoxy resin, polyurethane, etc.) and its curing conditions (curing temperature, curing time, etc.) are obtained from the process parameter layer. Based on the information of the encapsulation structure parameters and the process parameter layer, glue flow simulation and thermo-mechanical performance calculation are carried out. Glue flow simulation uses computational fluid dynamics technology to simulate the flow of the glue: a 3D geometric model is established and the physical properties and boundary conditions of the glue are defined, then mesh generation is carried out and the flow equation is solved, and data such as the flow velocity and pressure distribution of the glue are analyzed. By checking the uniformity of the glue distribution in each area, problems of uneven flow are found, and the dispensing path, injection parameters, or geometric design are adjusted according to the results to optimize the uniform distribution and fluidity of the glue and ensure the effectiveness of the final encapsulation process. Thermo-mechanical performance calculation evaluates the temperature distribution, thermal expansion, and stress changes during the glue curing process through thermodynamic simulation and mechanical stress analysis: based on the curing conditions, thermal analysis is carried out to calculate the temperature field during the glue curing process to ensure uniform temperature distribution. Then, the thermal expansion difference between the glue and the encapsulation material is evaluated through thermal expansion analysis, and the stress distribution generated during the curing process is calculated to avoid encapsulation deformation or cracks caused by thermal stress. Finally, through finite element analysis, thermal stress and deformation are comprehensively considered to optimize the process parameters and ensure the stability and reliability of the encapsulation structure. Then, according to the results of the glue flow simulation and the calculation of the thermo-mechanical performance, the overall performance of the preselected lamination scheme is evaluated, evaluating whether the glue can be evenly distributed between the layers of the lamination structure to avoid encapsulation failure caused by uneven glue distribution. The uniformity of the glue distribution directly affects the bonding effect and overall strength of the encapsulation. The mechanical stress generated due to thermal expansion and temperature changes during the curing process is evaluated to ensure that the lamination structure can remain stable after curing and will not cause encapsulation failure due to excessive thermal stress. According to these evaluation results, a lamination scheme feedback is generated.

[0035] Furthermore, select the corresponding preselected lamination scheme as the optimal lamination scheme according to the lamination scheme feedback, including: Generate a call score according to the glue distribution characteristics and thermo-mechanical performance; Set a performance evaluation threshold according to the historical lamination scheme. Compare the call score with the performance evaluation threshold. If the call score exceeds the performance evaluation threshold, mark the corresponding preselected lamination scheme as a lamination candidate scheme. Sort several lamination candidate schemes to obtain the optimal lamination scheme.

[0036] Specifically, according to the comprehensive evaluation of the glue distribution characteristics and thermo-mechanical properties, generate the glue distribution score and thermo-mechanical property score for each preselected lamination scheme. The call score can be generated using the weighted average method, which weights the scores of the glue distribution characteristics and thermo-mechanical properties and averages them. The weights of the two are set according to the importance of the process requirements. For example, if the uniformity of the glue is more important than the thermo-mechanical properties for a certain encapsulation process, a higher weight is given to the glue distribution characteristics. The calculation formula for the call score is: ; Where, and are the weight coefficients of the glue distribution characteristics and thermo-mechanical properties respectively. Then, according to the performance data of the historical lamination scheme, set a performance evaluation threshold, which is set by analyzing the performance data of the historical successful schemes and extracting them. Compare the call score of each preselected lamination scheme with the set performance evaluation threshold. If the call score exceeds the performance evaluation threshold, mark the preselected lamination scheme as a "lamination candidate scheme". This marking indicates that the scheme performs well in terms of glue distribution characteristics and thermo-mechanical properties, has strong adaptability, and meets the process requirements. Once multiple lamination candidate schemes are determined, sort the multiple lamination candidate schemes and obtain the optimal lamination scheme according to the sorting result.

[0037] Furthermore, sorting the lamination candidate schemes to obtain the optimal lamination scheme includes: S3: Mark any lamination candidate scheme as the sequence base point, compare the remaining lamination candidate schemes with the sequence base point respectively, and adjust the relative positions of the remaining lamination candidate schemes and the sequence base point according to the comparison results to obtain two unordered lamination sets. S4: Cancel the marking of the lamination candidate scheme, select any lamination candidate scheme in the unordered lamination set as the sequence base point, and compare it with the remaining lamination candidate schemes in the unordered lamination set, and adjust the relative positions respectively. Repeat steps S3 and S4 until several lamination candidate schemes are ordered to obtain the optimal lamination scheme.

[0038] As an optimization of the above embodiments, one of all the lamination candidate solutions is selected as the sequence base point, and the remaining lamination candidate solutions are compared with this base point one by one. By comparing the correlation and adaptability of each candidate solution with the base point, their relative positions are adjusted to form two unordered sets of laminations. At this time, all candidate solutions that are not selected as the base point will be regarded as elements in the unordered set; the comparison criteria can be glue distribution, thermo-mechanical properties, or other evaluation criteria. Through these criteria, the differences between each candidate solution and the sequence base point are judged, and their relative positions are adjusted according to the degree of difference. Through comparison and adjustment, the relative position of each candidate solution with the sequence base point will be optimized, and finally two separate unordered sets will be formed, one containing solutions that are relatively similar to the base point, and the other containing solutions that are quite different from the base point. Among them, when adjusting the relative position, the multi-dimensional evaluation results of each solution (such as compatibility score, process effect, production cost, etc.) are considered, and multi-attribute decision-making methods (such as Analytic Hierarchy Process AHP) can be combined to more scientifically adjust the sorting of candidate solutions. Then, the mark of the current sequence base point is cancelled, and a new solution in the unordered set of laminations is selected as the new sequence base point, and then the new base point is compared with the remaining candidate solutions and their relative positions are adjusted to form a new unordered set. In this way, the comparison and adjustment steps are repeatedly executed until all candidate solutions are sorted in an orderly manner according to their relative adaptability. When updating the base point, the new sequence base point is randomly called from the unordered set. With each adjustment, the relative positions of the candidate solutions are gradually optimized, and finally all candidate solutions will be sorted according to the evaluation criteria. After multiple rounds of iterative adjustment, all lamination candidate solutions will be ordered to obtain a sorting result, and the final optimal lamination solution is obtained according to the sorting result.

[0039] Embodiment 2; Based on the same inventive concept as an evaluation method for a multi-layer chip stacking scheme in the foregoing embodiments, the present invention also provides an evaluation system for a multi-layer chip stacking scheme, as Figure 5 shown. The system includes: A variable acquisition and integration module that determines the scheme variable items and collects data for each scheme variable item to obtain a number of variable information sets. The scheme variable items constitute the lamination scheme in the stacked packaging process; A process scheme generation module that selects the process information in the corresponding variable information set for each scheme variable item and obtains a number of basic process schemes; A process scheme optimization module that directionally adjusts the process information of the scheme variable items in the basic process scheme to obtain a number of optimized process schemes, and takes the number of optimized process schemes and the basic process schemes as preselected lamination schemes; A virtual simulation evaluation module that constructs a virtual process mirror image, digitally simulates the preselected lamination schemes respectively, and obtains lamination scheme feedback; The solution optimization decision-making module selects the corresponding preselected lamination solution as the best lamination solution according to the lamination solution feedback.

[0040] The above adjustment system in the present invention can effectively implement the evaluation method of the multi-layer chip stacking solution, and the technical effects that can be achieved are as described in the above embodiments, which will not be elaborated here.

[0041] Furthermore, the virtual simulation evaluation module includes: The virtual mirror construction unit extracts the packaging material parameters according to the preselected lamination solution and constructs a virtual process mirror, which is divided into a geometric structure layer and a process parameter layer; The parameter extraction and analysis unit obtains the packaging structure parameters according to the geometric structure layer, and obtains the glue type and curing conditions according to the process parameter layer; The fluid-thermal coupling simulation unit performs glue flow simulation according to the packaging structure parameters, glue type and curing conditions, obtains the glue distribution characteristics, and calculates the thermo-mechanical properties based on the process parameter layer; The solution performance evaluation unit evaluates the preselected lamination solution according to the glue distribution characteristics and thermo-mechanical properties, and obtains the lamination solution feedback.

[0042] Similarly, for the above optimization solutions of the system, the corresponding optimization effects of the method in Embodiment 1 can also be realized respectively, which will not be elaborated here either.

[0043] Although the present application has been described in conjunction with specific features and their embodiments, it is obvious that various modifications and combinations can be made without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely exemplary illustrations of the present application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A method for evaluating a multi-layer chip stacking solution, characterized in that: The method comprises: Determine solution variable items, and collect data for each of the solution variable items to obtain a plurality of variable information sets, wherein the solution variable items constitute a stacking solution in a stacking packaging process; Each of the scheme variable items selects the process information corresponding to the variable information set and obtains several basic process schemes; Directively adjusting the process information of the solution variable items in the basic process solution to obtain a number of optimized process solutions, and using the several optimized process solutions and the basic process solution as pre-selected lamination solutions; Construct a virtual process image, perform digital process simulation on the pre-selected stacking schemes, and obtain stacking scheme feedback; According to the lamination scheme feedback, the corresponding pre-selected lamination scheme is selected as the optimal lamination scheme.

2. The method for evaluating a multi-layer chip stacking solution according to claim 1, characterized in that: Each of the scheme variable items selects the process information corresponding to the variable information set and obtains several basic process schemes, including: Establishing a lamination process database, extracting historical lamination schemes according to the lamination process database, and marking the scheme variable items corresponding to the historical lamination schemes as pre-selected combination schemes, wherein the scheme variable items include glue type information, glue dispensing path information and lamination geometry information; Extracting a plurality of dispensing path information and lamination geometry information according to the lamination process database, and evaluating the compatibility of each dispensing path information with a plurality of lamination geometry information; Classifying the historical lamination schemes according to the glue type information to obtain a plurality of historical lamination scheme sets, wherein each of the historical lamination scheme sets corresponds to each of the glue type information one by one; The calling weight is set according to the compatibility evaluation result and the pre-selected combination scheme, and the process information of the remaining scheme variable items of each of the historical glue stacking scheme sets is reorganized according to the calling weight to obtain a number of basic process schemes.

3. The method for evaluating a multi-layer chip stacking solution according to claim 2, characterized in that: The compatibility of the plurality of dispensing path information and the plurality of lamination geometry information is evaluated respectively, including: Extracting interlayer geometric parameters according to the stacking geometric information, and determining geometric process thresholds based on packaging process requirements; Calculating the geometric adaptation range of the glue dispensing path information, obtaining the glue quantity distribution constraint according to the packaging process requirements, and screening the laminate geometric information based on the glue quantity distribution constraint; Detecting whether the projection area of ​​the dispensing path information in the stacking packaging direction and the edge spacing of the stacking geometric information meet a preset safety threshold, and obtaining a path conflict detection result according to the spacing detection result; According to the geometric adaptation range, geometric process threshold and path conflict detection result, a compatibility evaluation result of the dispensing path information and the stacking geometric information is generated.

4. The method for evaluating a multi-layer chip stacking solution according to claim 2, characterized in that: Directively adjust the process information of the solution variable items in the basic process solution to obtain several optimized process solutions, including: S1: extracting a number of recombinant gel stacking scheme sets according to the basic process scheme, performing information migration on the recombinant gel stacking scheme sets based on the compatibility evaluation result, marking schemes that are identical to the basic process scheme after exchange and screening them out, and obtaining exchange process schemes, wherein the information migration is the exchange of process information of corresponding scheme variable items between different recombinant gel stacking scheme sets; S2: Directively adjusting a single variable item in the exchange process scheme, and re-evaluating the compatibility of the exchange process scheme after the direct adjustment to obtain an optimized process scheme; The optimized process plan is used as the optimized basic process plan to repeat steps S1 and S2 until a preset number of iterations is reached to obtain a number of optimized process plans.

5. The method for evaluating a multi-layer chip stacking solution according to claim 2, characterized in that: The calling weight is set according to the compatibility evaluation result and the pre-selected combination scheme, including: Based on the compatibility evaluation result, generating a compatibility score for each of the preselected combination solutions; Extracting the process success rate of the historical lamination scheme according to the lamination process database, and assigning basic weights to the historical scheme variable items according to the process success rate; The compatibility score and the basic weight are linearly weighted to obtain a call weight.

6. The method for evaluating a multi-layer chip stacking solution according to claim 1, characterized in that: Performing process digital simulation on the pre-selected lamination schemes respectively to obtain lamination scheme feedback, including: Extracting packaging material parameters according to the preselected lamination scheme and constructing a virtual process image, wherein the virtual process image is divided into a geometric structure layer and a process parameter layer; Acquire packaging structure parameters according to the geometric structure layer, and acquire glue type and curing conditions according to the process parameter layer; Perform glue flow simulation according to the packaging structure parameters, glue type and curing conditions to obtain glue distribution characteristics, and calculate thermomechanical properties based on the process parameter layer; The preselected lamination scheme is evaluated according to the glue distribution characteristics and the thermomechanical properties to obtain lamination scheme feedback.

7. The method for evaluating a multi-layer chip stacking solution according to claim 6, characterized in that: Selecting the corresponding pre-selected lamination scheme as the best lamination scheme according to the lamination scheme feedback includes: generating a call score based on the glue distribution characteristics and the thermomechanical properties; A performance evaluation threshold is set according to the historical lamination scheme, and the call score is compared with the performance evaluation threshold. If the call score exceeds the performance evaluation threshold, the pre-selected lamination scheme is marked as a candidate lamination scheme; The candidate stacking solutions are sorted to obtain the best stacking solution.

8. The method for evaluating a multi-layer chip stacking solution according to claim 7, characterized in that: Sorting the candidate stacking solutions to obtain the best stacking solution includes: S3: marking any of the stacking candidate solutions as a sequence base point, comparing the remaining stacking candidate solutions with the sequence base point respectively, adjusting the relative positions of the remaining stacking candidate solutions and the sequence base point according to the comparison results, and obtaining two stacking unordered sets; S4: canceling the mark of the stacking candidate solution, and selecting any stacking candidate solution in the stacking unordered set as the sequence base point, and comparing it with the remaining stacking candidate solutions in the stacking unordered set, and adjusting the relative positions respectively; Repeat steps S3 and S4 until the stacking candidate solutions are sorted and the best stacking solution is obtained.

9. An evaluation system for a multi-layer chip stacking solution, characterized in that: The system comprises: A variable collection and integration module determines the scheme variable items and collects data for each of the scheme variable items to obtain a plurality of variable information sets, wherein the scheme variable items constitute a stacking scheme in a stacking packaging process; A process scheme generating module, wherein each of the scheme variable items selects the process information corresponding to the variable information set and obtains a number of basic process schemes; A process scheme tuning module is used to adjust the process information of the scheme variable items in the basic process scheme in a targeted manner, obtain a number of optimized process schemes, and use the several optimized process schemes and the basic process scheme as pre-selected lamination schemes; A virtual simulation evaluation module is used to construct a virtual process image, perform digital process simulation on the pre-selected stacking schemes, and obtain stacking scheme feedback; The scheme optimization decision module selects the corresponding pre-selected stacking scheme as the best stacking scheme according to the stacking scheme feedback.

10. The evaluation system for a multi-layer chip stacking solution according to claim 9, characterized in that: The virtual simulation evaluation module includes: A virtual image building unit, which extracts packaging material parameters according to the pre-selected lamination scheme and builds a virtual process image, wherein the virtual process image is divided into a geometric structure layer and a process parameter layer; A parameter extraction and analysis unit, which obtains packaging structure parameters according to the geometric structure layer, and obtains glue type and curing conditions according to the process parameter layer; A fluid-heat coupling simulation unit performs glue flow simulation according to the packaging structure parameters, glue type and curing conditions, obtains glue distribution characteristics, and calculates thermomechanical properties based on the process parameter layer; The scheme performance evaluation unit evaluates the pre-selected lamination scheme according to the glue distribution characteristics and the thermomechanical properties, and obtains feedback on the lamination scheme.

Images