Method, device and equipment for optimizing through silicon via array structure

Abstract

The invention provides a method, a device and equipment for optimizing a silicon through hole array structure. Acquiring preset key parameter factors corresponding to the system-level TSV array structure; uniformly generating a parameter level based on the preset key parameter factor and the parameter range of the preset key parameter factor; obtaining an equivalent stress value and an equivalent strain value corresponding to the parameter level by using a curved surface response method and a finite element simulation method through the parameter level; performing equation fitting on the equivalent stress value and the equivalent strain value by adopting a multiple regression equation fitting and variance analysis method to correspondingly obtain a final fitting stress regression equation and a final fitting strain regression equation; and according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the values of the preset key parameter factors are optimized by adopting a genetic algorithm, the stress optimal key parameter value and the strain optimal key parameter value are correspondingly obtained, and the accuracy of the optimal key parameter values is improved.

Description

Technical Field

[0001] The present invention relates to the field of microelectronic packaging technology and simulation optimization design technology, and in particular to an optimization method, device and equipment for a through silicon via array structure. Background Art

[0002] With the rapid development of modern electronic technology, electronic equipment is increasingly used in airborne, vehicle-mounted and mobile devices. These devices often need to work continuously in harsh environments such as high temperature, high humidity and vibration, so higher requirements are placed on the reliability of the integrated circuit (IC) packaging devices used in them. As one of the advanced packaging technologies, the structural reliability of TSV (Through Silicon Via) interconnection technology is directly related to the overall performance of the packaged device. However, in harsh environments or specific conditions, the reliability problems of TSV interconnection structures have gradually become prominent, such as stress concentration and increased fatigue loss. These problems directly or indirectly affect the reliability and life of the packaged device. Therefore, how to optimize the design of the TSV interconnection structure to improve its continuous working ability under conditions such as high temperature, high humidity and vibration has become a technical problem that needs to be solved urgently.

[0003] In order to solve the above problems, domestic and foreign scholars have conducted in-depth research on the reliability and structural optimization of TSV interconnection structures in recent years. The main research directions include reliability analysis in terms of heat, vibration, etc., as well as optimization design for single factor single target or multi-factor single target. In terms of research methods, the optimal combination of structural parameters is usually selected through multi-factor orthogonal experimental design. Orthogonal experimental design selects factors such as solder joint height, solder joint diameter, micro-bump height, micro-bump diameter, material, TSV array arrangement, etc. as independent variables, designs a series of test schemes under different structural parameters, and solves the stress and strain values ​​of each group of schemes in the simulation software. In addition, some studies also use the gray correlation method to further expand the sample capacity, and obtain the weight ratio of each parameter by analyzing the relationship between the independent variable and the target indicator, so as to obtain the optimal solution.

[0004] Although the prior art has made certain progress in the reliability and structural optimization of TSV interconnect structures, there are still many shortcomings. First of all, although the orthogonal experimental design can obtain the optimal level combination for a single objective, it is powerless for multi-objective optimization design problems. Taking the orthogonal optimization design of the TSV array interconnect structure as an example, even if only four parameter levels and four parameter factors are considered, a large number of simulation experiments are required to obtain comprehensive and accurate results. However, the prior art usually only conducts a limited number of simulation experiments, resulting in unknown indicator information for most of the remaining solutions. In addition, although the gray correlation analysis method can solve the multi-objective optimization design problem to a certain extent, due to the limited sample size, the value of the correlation degree is not accurate enough, and it is difficult to support higher-precision structural optimization goals. Therefore, how to provide a reliability analysis method for TSV interconnect structure with multi-objective optimization design has become a technical problem that the prior art urgently needs to solve. Summary of the invention

[0005] In order to solve the above problems existing in the prior art, the present invention provides a method, device and equipment for optimizing a through silicon via array structure.

[0006] The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0007] In a first aspect, the present invention provides a method for optimizing a through silicon via array structure, comprising:

[0008] Obtaining preset key parameter factors corresponding to the system-level TSV array structure;

[0009] Uniformly generating parameter levels based on preset key parameter factors and parameter ranges of preset key parameter factors;

[0010] Through the parameter level, the surface response method and finite element simulation method are used to obtain the equivalent stress value and equivalent strain value corresponding to the parameter level;

[0011] The multivariate regression equation fitting and variance analysis method were used to fit the equivalent stress value and the equivalent strain value, and the final fitting stress regression equation and the final fitting strain regression equation were obtained respectively;

[0012] According to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the genetic algorithm is used to optimize the values ​​of the preset key parameter factors respectively, and the optimal key parameter values ​​of stress and strain are obtained accordingly.

[0013] Optionally, the preset key parameter factors include: Cu column height, Cu column diameter, SiO2 layer thickness, TSV array spacing, redistribution layer RDL thickness and redistribution layer RDL diameter.

[0014] Optionally, uniformly generating parameter levels based on preset key parameter factors and parameter ranges of the preset key parameter factors includes:

[0015] Selecting a preset number of target key parameter factors from preset key parameter factors;

[0016] Within the parameter range of the target key parameter factor, uniform values ​​are taken according to the arithmetic distribution to obtain the parameter level corresponding to each target key parameter factor.

[0017] Optionally, the equivalent stress value and equivalent strain value corresponding to the parameter level are obtained by using the surface response method and the finite element simulation method through the parameter level, including:

[0018] Using parameter levels, the surface response method is used to perform permutations and combinations to obtain multiple parameter level combinations;

[0019] Based on the combination of multiple parameter levels, the finite element simulation method is used to perform stress simulation processing to obtain the corresponding equivalent stress value;

[0020] Based on the combination of multiple parameter levels, the finite element simulation method is used to perform strain simulation and obtain the corresponding equivalent strain value.

[0021] Optionally, multiple regression equation fitting and variance analysis methods are used to fit the equivalent stress value and the equivalent strain value respectively, and the final fitting stress regression equation and the final fitting strain regression equation are obtained accordingly, including:

[0022] Through the combination of equivalent stress values ​​and corresponding multiple parameter levels, the stress equation is fitted using the multivariate regression equation fitting method to obtain the initial fitting stress regression equation;

[0023] Through the combination of equivalent strain values ​​and corresponding multiple parameter levels, the strain equation is fitted using the multivariate regression equation fitting method to obtain the initial fitting strain regression equation;

[0024] The variance analysis method was used to verify the non-significance of the initial fitting stress regression equation and the initial fitting strain regression equation, and the non-significant items in the initial fitting stress regression equation and the initial fitting strain regression equation were deleted to obtain the final fitting stress regression equation and the final fitting strain regression equation respectively.

[0025] Optionally, the initial fitting stress regression equation is expressed as:

[0026]

[0027] Wherein, Y1 represents the equivalent stress value, β0 represents the stress reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, β1 represents the first linear stress regression coefficient of the first-order term, β2 represents the second linear stress regression coefficient of the first-order term, β3 represents the third linear stress regression coefficient of the first-order term, β4 represents the fourth linear stress regression coefficient of the first-order term, β 11 represents the first nonlinear stress regression coefficient of the quadratic term, β 22 The second nonlinear stress regression coefficient representing the quadratic term, β 33 represents the third nonlinear stress regression coefficient of the quadratic term, β 44 The fourth nonlinear stress regression coefficient of the quadratic term, β 12 The first stress regression coefficient, β, represents the interaction term 13 The second stress regression coefficient, β, represents the interaction term 14 The third stress regression coefficient, β 23 The fourth stress regression coefficient, β 24 The fifth stress regression coefficient, β, represents the interaction term 34 represents the sixth stress regression coefficient of the interaction term;

[0028] The initial fitting strain regression equation is expressed as:

[0029]

[0030] Wherein, Y2 represents the equivalent strain value, γ0 represents the strain reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, γ1 represents the first linear strain regression coefficient of the first-order term, γ2 represents the second linear strain regression coefficient of the first-order term, γ3 represents the third linear strain regression coefficient of the first-order term, γ4 represents the fourth linear strain regression coefficient of the first-order term, γ 11 represents the first nonlinear strain regression coefficient of the quadratic term, γ 22 The second nonlinear strain regression coefficient representing the quadratic term, γ 33 represents the third nonlinear strain regression coefficient of the quadratic term, γ 44 represents the fourth nonlinear strain regression coefficient of the quadratic term, γ 12 represents the first strain regression coefficient of the interaction term, γ 13 The second strain regression coefficient, γ, represents the interaction term 14 The third strain regression coefficient, γ, represents the interaction term 23 The fourth strain regression coefficient, γ, represents the interaction term 24 The fifth strain regression coefficient, γ, represents the interaction term34 Represents the sixth strain regression coefficient of the interaction term.

[0031] Optionally, according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the values ​​of the preset key parameter factors are optimized by using a genetic algorithm to obtain the optimal key parameter value of stress and the optimal key parameter value of strain, including:

[0032] A genetic algorithm is used to select a first parameter factor value within a parameter range of a preset key parameter factor, and the first parameter factor value is brought into a final fitting stress regression equation to obtain a plurality of equivalent stress fitting values;

[0033] Select the first parameter factor value corresponding to the minimum equivalent stress fitting value as the optimal key parameter value of stress;

[0034] A genetic algorithm is used to select a second parameter factor value within a parameter range of a preset key parameter factor, and the second parameter factor value is brought into a final fitting strain regression equation to obtain a plurality of equivalent strain fitting values;

[0035] The second parameter factor value corresponding to the minimum equivalent strain fitting value is selected as the optimal key parameter value of strain.

[0036] In a second aspect, the present invention provides an optimization device for a through silicon via array structure, the optimization device for a through silicon via array structure comprising: an acquisition unit, a parameter generation unit, a simulation unit, a fitting unit, and an optimization unit;

[0037] The acquisition unit is used to: acquire preset key parameter factors corresponding to the system-level TSV array structure;

[0038] The parameter generation unit is used to: uniformly generate parameter levels based on preset key parameter factors and parameter ranges of the preset key parameter factors;

[0039] The simulation unit is used to: obtain the equivalent stress value and equivalent strain value corresponding to the parameter level by using the surface response method and the finite element simulation method through the parameter level;

[0040] The fitting unit is used to: adopt multiple regression equation fitting and variance analysis method to respectively fit the equivalent stress value and the equivalent strain value, and obtain the final fitting stress regression equation and the final fitting strain regression equation accordingly;

[0041] The optimization unit is used to optimize the values ​​of the preset key parameter factors using a genetic algorithm according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, and obtain the optimal key parameter value of stress and the optimal key parameter value of strain accordingly.

[0042] In a third aspect, the present invention provides an optimization device for a through silicon via array structure, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor. When the optimization device for the through silicon via array structure is running, the processor and the storage medium communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of the optimization method for a through silicon via array structure as described in any one of the first aspects above.

[0043] The present invention provides a method, device and equipment for optimizing a through silicon via array structure. Among them, a method for optimizing a through silicon via array structure includes: obtaining preset key parameter factors corresponding to a system-level TSV array structure; uniformly generating parameter levels based on the preset key parameter factors and the parameter range of the preset key parameter factors; obtaining the equivalent stress value and the equivalent strain value corresponding to the parameter level by using the surface response method and the finite element simulation method through the parameter level; using the multivariate regression equation fitting and variance analysis method to respectively perform equation fitting on the equivalent stress value and the equivalent strain value, and correspondingly obtain the final fitting stress regression equation and the final fitting strain regression equation; according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, using the genetic algorithm to respectively optimize the values ​​of the preset key parameter factors, and correspondingly obtain the optimal key parameter value of stress and the optimal key parameter value of strain. In the present invention, by adopting a method combining the surface response method with the genetic algorithm to perform multi-factor and multi-objective optimization design, it is possible to traverse more comprehensive parameter combinations and find potential better solutions, so that the final optimal key parameter values ​​are more accurate; secondly, based on the parameter optimization method of the present invention, the optimal key parameter values ​​of stress and the optimal key parameter values ​​of strain can be optimized at the same time, and multiple objectives can be optimized at the same time, thereby improving the comprehensive reliability of the TSV interconnect structure parameter optimization results.

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic flow chart of a method for optimizing a through silicon via array structure provided by an embodiment of the present invention;

[0046] Figure 2 A schematic structural diagram of a through silicon via array structure optimization device provided by an embodiment of the present invention;

[0047] Figure 3 A schematic structural diagram of a through silicon via array structure optimization device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0049] In order to improve the accuracy of the obtained optimal key parameter values ​​and improve the comprehensive reliability of the optimization results, an embodiment of the present invention provides an optimization method for a through silicon via array structure. Figure 1 A schematic diagram of a process flow of a through silicon via array structure optimization method provided by an embodiment of the present invention. Figure 1 As shown, the optimization method includes:

[0050] S101 : Obtain preset key parameter factors corresponding to a system-level TSV array structure.

[0051] Optionally, the preset key parameter factors include: Cu column height, Cu column diameter, SiO2 layer thickness, TSV array spacing, redistribution layer RDL thickness and redistribution layer RDL diameter.

[0052] It should be noted that the TSV (Through-Silicon Via) array structure is a key component in three-dimensional integrated circuit (3D-IC) technology. It allows vertical connection of chips or wafers at different levels, thereby achieving higher integration, smaller size, lower power consumption and faster data transmission speed. In the system-level TSV array structure, multiple TSVs are organized and arranged in a specific layout to meet the requirements of electrical performance, thermal management, mechanical strength and other aspects.

[0053] Furthermore, in some possible implementations, the preset key parameter factors may also include: TSV resistance, interconnect capacitance, thermal conductivity, adhesion between TSV and silicon interface, number of RDL layers, lithography resolution, deposition rate and uniformity, fatigue life, solder joint reliability, environmental tolerance and design margin, etc. The setting of the preset key parameter factors can be flexibly set according to actual needs, and this embodiment does not limit this.

[0054] S102: uniformly generate parameter levels based on preset key parameter factors and parameter ranges of the preset key parameter factors.

[0055] It should be noted that, since the method of the present invention studies the continuous working ability of the TSV structure under conditions of high temperature, high humidity and vibration, in this embodiment, the parameter range of the preset key parameter factor can be the parameter range of the preset key parameter factor corresponding to the TSV structure under conditions of high temperature, high humidity and vibration, and the parameter range can be specifically determined by empirical data or in the laboratory.

[0056] Optionally, S102 may specifically include:

[0057] Selecting a preset number of target key parameter factors from preset key parameter factors;

[0058] Within the parameter range of the target key parameter factor, uniform values ​​are taken according to the arithmetic distribution to obtain the parameter level corresponding to each target key parameter factor.

[0059] In this embodiment, a preset number of target key parameter factors of interest can be selected from the preset key parameter factors according to the optimization requirements, and the parameter level corresponding to each target key parameter factor can be uniformly selected within the parameter range of the target key parameter factor. For example, when the parameter range corresponding to the Cu column height is a1≤A≤a2, when 5 parameter levels need to be selected, the parameter levels correspond to: a1, a2.

[0060] S103. By using the parameter level, the surface response method and the finite element simulation method are used to obtain the equivalent stress value and the equivalent strain value corresponding to the parameter level.

[0061] Optionally, S103 may specifically include:

[0062] Using parameter levels, the surface response method is used to perform permutations and combinations to obtain multiple parameter level combinations;

[0063] Based on the combination of multiple parameter levels, the finite element simulation method is used to perform stress simulation processing to obtain the corresponding equivalent stress value;

[0064] Based on the combination of multiple parameter levels, the finite element simulation method is used to perform strain simulation and obtain the corresponding equivalent strain value.

[0065] S104. Using multiple regression equation fitting and variance analysis methods, the equivalent stress value and the equivalent strain value are respectively fitted with equations to obtain the final fitting stress regression equation and the final fitting strain regression equation.

[0066] Optionally, S104 may specifically include:

[0067] Through the combination of equivalent stress values ​​and corresponding multiple parameter levels, the stress equation is fitted using the multivariate regression equation fitting method to obtain the initial fitting stress regression equation;

[0068] Through the combination of equivalent strain values ​​and corresponding multiple parameter levels, the strain equation is fitted using the multivariate regression equation fitting method to obtain the initial fitting strain regression equation;

[0069] The variance analysis method was used to verify the non-significance of the initial fitting stress regression equation and the initial fitting strain regression equation, and the non-significant items in the initial fitting stress regression equation and the initial fitting strain regression equation were deleted to obtain the final fitting stress regression equation and the final fitting strain regression equation respectively.

[0070] Optionally, the initial fitting stress regression equation is expressed as:

[0071]

[0072] Wherein, Y1 represents the equivalent stress value, β0 represents the stress reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, β1 represents the first linear stress regression coefficient of the first-order term, β2 represents the second linear stress regression coefficient of the first-order term, β3 represents the third linear stress regression coefficient of the first-order term, β4 represents the fourth linear stress regression coefficient of the first-order term, β 11 represents the first nonlinear stress regression coefficient of the quadratic term, β 22 The second nonlinear stress regression coefficient representing the quadratic term, β 33 represents the third nonlinear stress regression coefficient of the quadratic term, β 44 The fourth nonlinear stress regression coefficient of the quadratic term, β 12 The first stress regression coefficient, β, represents the interaction term 13 The second stress regression coefficient, β, represents the interaction term 14 The third stress regression coefficient, β 23 The fourth stress regression coefficient, β 24 The fifth stress regression coefficient, β, represents the interaction term 34 represents the sixth stress regression coefficient of the interaction term;

[0073] The initial fitting strain regression equation is expressed as:

[0074]

[0075] Wherein, Y2 represents the equivalent strain value, γ0 represents the strain reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, γ1 represents the first linear strain regression coefficient of the first-order term, γ2 represents the second linear strain regression coefficient of the first-order term, γ3 represents the third linear strain regression coefficient of the first-order term, γ4 represents the fourth linear strain regression coefficient of the first-order term, γ 11 represents the first nonlinear strain regression coefficient of the quadratic term, γ 22 The second nonlinear strain regression coefficient representing the quadratic term, γ 33 represents the third nonlinear strain regression coefficient of the quadratic term, γ 44 represents the fourth nonlinear strain regression coefficient of the quadratic term, γ 12 represents the first strain regression coefficient of the interaction term, γ 13 The second strain regression coefficient, γ, represents the interaction term 14 The third strain regression coefficient, γ, represents the interaction term23 The fourth strain regression coefficient, γ, represents the interaction term 24 The fifth strain regression coefficient, γ, represents the interaction term 34 Represents the sixth strain regression coefficient of the interaction term.

[0076] Furthermore, the process of using the initial fitting stress regression equation and the initial fitting strain regression equation to obtain the final fitting stress regression equation or the final fitting strain regression equation through significance verification and variance analysis is specifically as follows:

[0077] First, use statistical software to calculate the variance analysis table, which shows the contribution (i.e., significance) of each parameter to the initial fitting stress regression equation and the initial fitting strain regression equation (hereinafter collectively referred to as the initial equation). Then, according to the set significance level (such as 0.05), determine which parameters are significant in the initial equation (i.e., have a significant impact on the initial equation).

[0078] Next, identify and delete non-significant terms, identify those non-significant parameter terms based on the variance analysis table, and update the initial equation, delete these non-significant terms from the initial equation, and obtain the updated equation.

[0079] In addition, refitting and verification can be performed to refit and verify the updated equation to ensure the stability and accuracy of the updated equation. The form or parameters of the updated equation can be further adjusted by refitting to optimize the fitting effect. Then the final fitting stress regression equation and the final fitting strain regression equation are obtained. The above final equation only contains those parameter items that have a significant impact on the equivalent stress value or the equivalent strain value, which ensures the accuracy and simplification effect of the final equation.

[0080] S105. According to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the values ​​of the preset key parameter factors are optimized by using a genetic algorithm to obtain the optimal key parameter value of stress and the optimal key parameter value of strain.

[0081] Optionally, S105 may specifically include:

[0082] A genetic algorithm is used to select a first parameter factor value within a parameter range of a preset key parameter factor, and the first parameter factor value is brought into a final fitting stress regression equation to obtain a plurality of equivalent stress fitting values;

[0083] Select the first parameter factor value corresponding to the minimum equivalent stress fitting value as the optimal key parameter value of stress;

[0084] A genetic algorithm is used to select a second parameter factor value within a parameter range of a preset key parameter factor, and the second parameter factor value is brought into a final fitting strain regression equation to obtain a plurality of equivalent strain fitting values;

[0085] The second parameter factor value corresponding to the minimum equivalent strain fitting value is selected as the optimal key parameter value of strain.

[0086] The embodiment of the present invention provides an optimization method for a through silicon via array structure. By adopting a method combining a surface response method with a genetic algorithm to perform multi-factor and multi-objective optimization design, it is possible to traverse more comprehensive parameter combinations and find potential better solutions, so that the final optimal key parameter values ​​are more accurate; secondly, with the goal of minimizing the equivalent stress fitting value and the equivalent strain fitting value, the optimal solution of the parameters is solved by a multi-objective optimization algorithm, which can optimize multiple objectives at the same time, thereby improving the comprehensive reliability of the TSV interconnect structure in temperature and vibration environments.

[0087] The method provided in the embodiment of the present invention can be applied to an electronic device. Specifically, the electronic device can be: a desktop computer, a portable computer, an intelligent mobile terminal, a server, etc., which is not limited in the embodiment of the present invention.

[0088] Based on the same inventive concept, an embodiment of the present invention further provides a device for optimizing a through silicon via array structure. Figure 2 A schematic diagram of a structure of an optimization device for a through silicon via array structure provided by an embodiment of the present invention. Figure 2 As shown, it includes: an acquisition unit 201, a parameter generation unit 202, a simulation unit 203, a fitting unit 204 and an optimization unit 205;

[0089] The acquisition unit 201 is used to: acquire preset key parameter factors corresponding to the system-level TSV array structure;

[0090] The parameter generating unit 202 is used to: uniformly generate parameter levels based on preset key parameter factors and parameter ranges of the preset key parameter factors;

[0091] The simulation unit 203 is used to obtain the equivalent stress value and the equivalent strain value corresponding to the parameter level by using the surface response method and the finite element simulation method according to the parameter level;

[0092] The fitting unit 204 is used to: adopt multiple regression equation fitting and variance analysis method to respectively fit the equivalent stress value and the equivalent strain value, and obtain the final fitting stress regression equation and the final fitting strain regression equation accordingly;

[0093] The optimization unit 205 is used to optimize the values ​​of the preset key parameter factors using a genetic algorithm according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, and obtain the optimal key parameter value of stress and the optimal key parameter value of strain.

[0094] Figure 3 A schematic diagram of a structure of a through silicon via array structure optimization device provided by an embodiment of the present invention includes: a processor 310, a storage medium 320 and a bus 330, wherein the storage medium 320 stores machine-readable instructions executable by the processor 310, and when the through silicon via array structure optimization device is running, the processor 310 communicates with the storage medium 320 via the bus 330, and the processor 310 executes the machine-readable instructions to execute the steps of the above method embodiment. The specific implementation method and technical effect are similar and will not be repeated here.

[0095] The storage medium may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk storage. Optionally, the storage medium may also be at least one storage device located away from the aforementioned processor.

[0096] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0097] It should be noted that the terms "first", "second", etc. are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention.

[0098] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification.

[0099] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art can understand and implement other changes of the above disclosed embodiments by viewing the drawings and the disclosed content. In the description of the present invention, the term "comprising" does not exclude other components or steps, "one" or "an" does not exclude multiple situations, and the meaning of "multiple" is two or more, unless otherwise clearly and specifically limited. In addition, certain measures are recorded in different embodiments, but this does not mean that these measures cannot be combined to produce good results.

[0100] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. A method for optimizing a through silicon via array structure, characterized in that: include: Obtaining preset key parameter factors corresponding to the system-level TSV array structure; Uniformly generating parameter levels based on the preset key parameter factors and parameter ranges of the preset key parameter factors; Through the parameter level, the equivalent stress value and the equivalent strain value corresponding to the parameter level are obtained by using the surface response method and the finite element simulation method; The equivalent stress value and the equivalent strain value are respectively fitted by multiple regression equation fitting and variance analysis method, and a final fitting stress regression equation and a final fitting strain regression equation are correspondingly obtained; According to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the values ​​of the preset key parameter factors are optimized respectively by using a genetic algorithm to obtain the optimal key parameter value of stress and the optimal key parameter value of strain.

2. The method for optimizing a through silicon via array structure according to claim 1, characterized in that: The preset key parameter factors include: Cu column height, Cu column diameter, SiO2 layer thickness, TSV array spacing, redistribution layer RDL thickness and redistribution layer RDL diameter.

3. The method for optimizing a through silicon via array structure according to claim 1, characterized in that: The uniformly generating parameter levels based on the preset key parameter factors and the parameter ranges of the preset key parameter factors comprises: Selecting a preset number of target key parameter factors from the preset key parameter factors; Within the parameter range of the target key parameter factor, values ​​are uniformly taken according to an arithmetic distribution to obtain the parameter level corresponding to each target key parameter factor.

4. The method for optimizing a through silicon via array structure according to claim 1, characterized in that: The method of obtaining the equivalent stress value and the equivalent strain value corresponding to the parameter level by using the surface response method and the finite element simulation method through the parameter level includes: Using the parameter levels, a surface response method is used to perform permutations and combinations to obtain multiple parameter level combinations; Based on the combination of the multiple parameter levels, a finite element simulation method is used to perform stress simulation processing to obtain the corresponding equivalent stress value; Based on the combination of the multiple parameter levels, a finite element simulation method is used to perform strain simulation processing to obtain the corresponding equivalent strain value.

5. The method for optimizing the through silicon via array structure according to claim 4, characterized in that: The multivariate regression equation fitting and variance analysis method are used to respectively fit the equivalent stress value and the equivalent strain value, and the final fitting stress regression equation and the final fitting strain regression equation are obtained accordingly, including: By combining the equivalent stress value and the corresponding multiple parameter levels, a stress equation is fitted using a multiple regression equation fitting method to obtain an initial fitting stress regression equation; By combining the equivalent strain value and the corresponding multiple parameter levels, a strain equation is fitted using a multiple regression equation fitting method to obtain an initial fitting strain regression equation; The initial fitting stress regression equation and the initial fitting strain regression equation are respectively verified to be non-significant by using the variance analysis method, and the non-significant terms in the initial fitting stress regression equation and the initial fitting strain regression equation are deleted to obtain the final fitting stress regression equation and the final fitting strain regression equation accordingly.

6. The method for optimizing the through silicon via array structure according to claim 5, characterized in that: The initial fitting stress regression equation is expressed as: Wherein, Y1 represents the equivalent stress value, β0 represents the stress reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, β1 represents the first linear stress regression coefficient of the first-order term, β2 represents the second linear stress regression coefficient of the first-order term, β3 represents the third linear stress regression coefficient of the first-order term, β4 represents the fourth linear stress regression coefficient of the first-order term, β 11 represents the first nonlinear stress regression coefficient of the quadratic term, β 22 The second nonlinear stress regression coefficient representing the quadratic term, β 33 represents the third nonlinear stress regression coefficient of the quadratic term, β 44 The fourth nonlinear stress regression coefficient of the quadratic term, β 12 The first stress regression coefficient, β, represents the interaction term 13 The second stress regression coefficient, β, represents the interaction term 14 The third stress regression coefficient, β 23 The fourth stress regression coefficient, β 24 The fifth stress regression coefficient, β, represents the interaction term 34 represents the sixth stress regression coefficient of the interaction term; The initial fitting strain regression equation is expressed as: Wherein, Y2 represents the equivalent strain value, γ0 represents the strain reference value, A represents the first target key parameter factor, B represents the second target key parameter factor, C represents the third target key parameter factor, D represents the fourth target key parameter factor, γ1 represents the first linear strain regression coefficient of the first-order term, γ2 represents the second linear strain regression coefficient of the first-order term, γ3 represents the third linear strain regression coefficient of the first-order term, γ4 represents the fourth linear strain regression coefficient of the first-order term, γ 11 represents the first nonlinear strain regression coefficient of the quadratic term, γ 22 The second nonlinear strain regression coefficient representing the quadratic term, γ 33 represents the third nonlinear strain regression coefficient of the quadratic term, γ 44 represents the fourth nonlinear strain regression coefficient of the quadratic term, γ 12 represents the first strain regression coefficient of the interaction term, γ 13 The second strain regression coefficient, γ, represents the interaction term 14 The third strain regression coefficient, γ, represents the interaction term 23 The fourth strain regression coefficient, γ, represents the interaction term 24 The fifth strain regression coefficient, γ, represents the interaction term 34 Represents the sixth strain regression coefficient of the interaction term.

7. The method for optimizing a through silicon via array structure according to claim 1, characterized in that: According to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, the values ​​of the preset key parameter factors are optimized by using a genetic algorithm to obtain the optimal key parameter value of stress and the optimal key parameter value of strain, including: A genetic algorithm is used to select a first parameter factor value within the parameter range of the preset key parameter factor, and the first parameter factor value is brought into the final fitting stress regression equation to obtain a plurality of equivalent stress fitting values; Selecting the first parameter factor value corresponding to the minimum equivalent stress fitting value as the optimal key parameter value of stress; A genetic algorithm is used to select a second parameter factor value within the parameter range of the preset key parameter factor, and the second parameter factor value is brought into the final fitting strain regression equation to obtain a plurality of equivalent strain fitting values; The second parameter factor value corresponding to the minimum equivalent strain fitting value is selected as the strain optimal key parameter value.

8. An optimization device for a through silicon via array structure, characterized in that: The optimization device for the through silicon via array structure comprises: an acquisition unit, a parameter generation unit, a simulation unit, a fitting unit and an optimization unit; The acquisition unit is used to: acquire preset key parameter factors corresponding to the system-level TSV array structure; The parameter generating unit is used to: uniformly generate parameter levels based on the preset key parameter factors and parameter ranges of the preset key parameter factors; The simulation unit is used to: obtain the equivalent stress value and the equivalent strain value corresponding to the parameter level by using the surface response method and the finite element simulation method through the parameter level; The fitting unit is used to: adopt multiple regression equation fitting and variance analysis method to respectively perform equation fitting on the equivalent stress value and the equivalent strain value, and obtain a final fitting stress regression equation and a final fitting strain regression equation accordingly; The optimization unit is used to: according to the final fitting stress regression equation, the final fitting strain regression equation and the parameter range of the preset key parameter factors, use a genetic algorithm to optimize the values ​​of the preset key parameter factors respectively, and correspondingly obtain the optimal key parameter value of stress and the optimal key parameter value of strain.

9. An optimization device for a through silicon via array structure, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the optimization device for the through silicon via array structure is running, the processor communicates with the storage medium through the bus, and the processor executes the machine-readable instructions to perform the steps of the optimization method for the through silicon via array structure as described in any one of claims 1 to 7.

Images