Mask pattern deviation value determination method and device, equipment, medium and product

Abstract

The invention provides a mask pattern deviation value determination method and device, equipment, a medium and a product, and the method comprises the steps: obtaining a pattern density distribution diagram of a historical design layout and pattern information of each pattern in the historical design layout, and the pattern information comprises a position relation between the patterns and a pattern type corresponding to each pattern; determining a long-range deviation evaluation model according to the graph density distribution diagram; determining a short-range deviation evaluation model according to the graphic information; and determining a deviation value of the mask pattern to be corrected according to the long-range deviation evaluation model and the short-range deviation evaluation model. According to the embodiment of the invention, the accuracy of determining the mask pattern deviation value can be improved.

Description

Technical Field

[0001] The present application relates to the field of semiconductor technology, and in particular to a method, device, equipment, medium and product for determining a mask pattern deviation value. Background Art

[0002] In the field of semiconductor manufacturing, as process nodes continue to shrink, the complexity and challenges of chip manufacturing are increasing. Especially in the pursuit of high yield, improving chip uniformity has become a crucial link. The existing mask pattern correction method based on the rule table mainly predicts the deviation value of the mask pattern through the preset pattern type and spacing data recorded in the rule table. However, the rule table only covers some preset specific patterns that may cause loading deviations. This limitation causes a large number of unselected patterns to still be affected by loading in actual production, resulting in inaccurate mask pattern deviation values ​​determined by the existing technology based on the rule table. Summary of the invention

[0003] The present application provides a method, device, equipment, medium and product for determining a mask pattern deviation value, which can improve the accuracy of determining a mask pattern deviation value.

[0004] In a first aspect, an embodiment of the present application provides a method for determining a mask pattern deviation value, the method comprising:

[0005] Obtaining a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout, wherein the graphic information includes a positional relationship between each graphic and a graphic type corresponding to each graphic;

[0006] Determining a long-range deviation assessment model based on the graphic density distribution map;

[0007] Determining a short-range deviation assessment model based on the graphical information;

[0008] The deviation value of the mask pattern to be corrected is determined according to the long-range deviation evaluation model and the short-range deviation evaluation model.

[0009] In a second aspect, the present application provides a device for determining a mask pattern deviation value, the device comprising:

[0010] An acquisition module, used to acquire a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout, wherein the graphic information includes a positional relationship between each graphic and a graphic type corresponding to each graphic;

[0011] A first determination module is used to determine a long-range deviation assessment model according to the graphic density distribution map;

[0012] A second determination module is used to determine a short-range deviation assessment model according to the graphic information;

[0013] The third determination module is used to determine the deviation value of the mask pattern to be corrected according to the long-range deviation evaluation model and the short-range deviation evaluation model.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising: a processor and a memory storing computer program instructions;

[0015] When the processor executes the computer program instructions, the method for determining the mask pattern deviation value in any one of the embodiments of the first aspect is implemented.

[0016] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, a method for determining a mask graphic deviation value as in any one of the embodiments in the first aspect is implemented.

[0017] In a fifth aspect, an embodiment of the present application provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes a method for determining a mask graphic deviation value as in any one of the embodiments in the first aspect above.

[0018] In a method, device, equipment, medium and product for determining a mask graphic deviation value provided in an embodiment of the present application, firstly, the detailed graphic information of each graphic in the historical design layout is obtained, which includes but is not limited to the precise positional relationship between the graphics and the corresponding graphic types. This information constitutes the basis of the short-range deviation evaluation model, which can capture the subtle deviations caused by the interaction between local patterns and make up for the shortcomings of the rule table in covering specific patterns. Furthermore, by parsing the graphic density distribution map of the historical design layout, the method constructs a long-range deviation evaluation model, which can evaluate the global deviation trend caused by the overall layout density difference, which is an important aspect ignored by the traditional rule table-based method. The combination of long-range and short-range evaluation models not only pays attention to the mutual influence between a single or a few graphics, but also examines the potential contribution of the entire design layout to the mask graphic deviation from a macro perspective, thereby realizing a multi-level and high-precision prediction of the deviation value, thereby improving the accuracy of the determination of the mask graphic deviation value. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0020] Figure 1 This is one of the flowcharts of the method for determining the deviation value of a mask pattern provided in an embodiment of the present application;

[0021] Figure 2 This is the second flow chart of the method for determining the mask pattern deviation value provided in the embodiment of the present application;

[0022] Figure 3 This is the third flow chart of the method for determining the mask pattern deviation value provided in the embodiment of the present application;

[0023] Figure 4 It is a schematic diagram of the principle of the method for determining the deviation value of a mask pattern provided in an embodiment of the present application;

[0024] Figure 5 It is a structural schematic diagram of a mask pattern deviation value determination device provided in an embodiment of the present application;

[0025] Figure 6 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0028] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0029] With the rapid development of chip manufacturing technology, the continuous miniaturization of key dimensions has put forward higher requirements for the precise control of process parameters, resulting in an increasingly narrow process window and a significant increase in yield loss factors. In order to improve the commercial value of chips, the industry generally focuses on yield improvement solutions, especially in the yield climbing stage, random defect reduction and uniformity improvement have become two core problems that need to be solved urgently. Random defect reduction involves complex cleanliness and purity control, and the process is long and full of variables; and although uniformity improvement can be fully detected and locate the root cause of the problem through advanced measurement equipment, the existing improvement methods are complicated to operate and difficult to deal with effectively.

[0030] The deterioration of uniformity is mainly attributed to the loading effect, which is a common phenomenon in various manufacturing processes such as etching and chemical mechanical polishing. Its impact is not limited to between chips, but also extends deeper into the chip. The optimization of the loading effect between chips depends on the efforts of the process end. However, the control accuracy of existing equipment is limited, and the cleaning challenge is huge; while the loading within the chip is mainly based on the rule-based compensation through the selective size amplification table (SSA table). Although this method can alleviate the problem to a certain extent, its limitations are also significant. The SSA table is not accurate enough to capture the long-distance loading effect and the loading differences of the same pattern under environmental changes, resulting in limited compensation effect.

[0031] Although the existing rule-based loading bias compensation process has formed a relatively fixed operation mode, including placing a specific pattern on the test mask, collecting data after development and etching, summarizing the deviation and generating the SSA table, and finally applying it to the product mask for manual compensation, its limitations are that it cannot fully reflect the loading status of the entire chip level and lacks sensitivity to the loading changes of the same pattern under different environments. In addition, the compensation accuracy of the SSA table is limited by its single-dimensional consideration based on pattern type and spacing data, and fails to fully incorporate the multiple factors generated by actual loading, so the compensation effect is often not ideal.

[0032] In order to solve the problems existing in the prior art, the embodiments of the present application provide a method, device, equipment, medium and product for determining a mask pattern deviation value.

[0033] The present application embodiment provides a method, device, equipment, medium and product for determining a mask pattern deviation value. The following first introduces the method for determining a mask pattern deviation value provided by the present application embodiment. Figure 1 As shown, the method specifically comprises the following steps:

[0034] S100, obtaining a graphic density distribution diagram of a historical design layout and graphic information of each graphic in the historical design layout, wherein the graphic information includes a positional relationship between each graphic and a graphic type corresponding to each graphic.

[0035] Optionally, in an embodiment of the present application, a graphic density distribution map is a visualization tool for showing the density of graphic elements in a design layout. Graphic density distribution maps of different ranges refer to dividing the entire design layout into multiple regions (or size ranges) and calculating the density of graphic elements in each region (i.e., the proportion of the graphic area to the total area of ​​the region). These regions can be grids divided according to fixed sizes, or they can be naturally formed regions based on the distribution of graphics. By generating a graphic density distribution map, the distribution of graphic density in the design layout can be intuitively seen, and then its impact on the manufacturing process and final product performance can be analyzed.

[0036] Optionally, the graphic information refers to a detailed description of each graphic element in the design layout. The positional relationship between the graphics refers to the relative positions between different graphic elements in the design layout. This includes the horizontal and vertical distances between them, and whether there is overlap or intersection. The accuracy of the positional relationship is crucial to ensure the correctness of the circuit function. For example, the metal line needs to be precisely aligned with the contact hole to ensure that the current can pass smoothly.

[0037] Graphic type refers to the circuit component or functional unit represented by each graphic element in the design layout. Different types of graphics play different roles in the circuit. For example, metal wires are used to connect different parts of the circuit, polysilicon gates are key components of transistors, and contact holes are used to achieve electrical connections between different layers. Accurately identifying the type of each graphic is crucial to understanding the overall structure and function of the design layout.

[0038] Optionally, in a feasible implementation of the present application, first, a data set containing historical design layouts is imported or accessed. These data sets may exist in the form of graphic files (such as GDSII, OASIS, etc.), which describe the layout design of the chip or circuit in detail.

[0039] Next, these graphic data are parsed using image processing and analysis technology to identify and extract the various graphic elements (such as lines, polygons, etc.) in the design layout. In this process, the position coordinates of each graphic, including its boundaries, center points, etc., can be recorded to construct a position relationship diagram between the graphics. At the same time, the types of these graphics can also be identified and classified, such as metal connections, polysilicon gates, contact holes, etc. This step is the key to understanding the functionality and structural characteristics of the design layout.

[0040] Subsequently, in order to obtain the graphic density distribution map, the entire design layout can be divided into regions, which can be based on the preset grid size or the natural distribution of the graphics. In each divided area, the ratio of the total area of ​​the graphic elements to the total area of ​​the area is calculated and counted, and this ratio represents the graphic density of the area. By traversing all areas and repeating the above calculation process, a graphic density distribution map representing the distribution of graphic density in different areas can be generated. This graphic density distribution map not only shows the spatial variation of graphic density in the design layout, but also provides an important basis for the subsequent analysis of the relationship between graphic density and mask graphic deviation.

[0041] Through this step, the graphic layout, graphic type and density distribution of the historical design layout can be fully and accurately understood, laying a solid data foundation for the subsequent construction of the mask graphic deviation evaluation model.

[0042] S200, determining a long-range deviation assessment model according to the graphic density distribution map.

[0043] Alternatively, long-range deviation refers to the deviation observed on the micrometer (μm) to millimeter (mm) scale. This deviation is mainly related to the change in pattern density. Pattern areas of different densities will affect the transport efficiency of reaction gases and reaction products during the etching process. Specifically, high-density areas may have a lower etching rate due to limited gas diffusion, while low-density areas may have a higher etching rate due to sufficient gas supply. This inconsistent etching rate caused by differences in pattern density will eventually manifest as a long-range deviation, that is, the deviation in etching size or shape observed between areas of different density.

[0044] Optionally, in a feasible implementation of the present application, the graphic density distribution map can be parsed using statistical methods or machine learning algorithms. Specifically including the statistical characteristics of the calculated density distribution, such as the mean value, standard deviation, peak value, valley value and the autocorrelation of the density distribution, these characteristics can reveal the overall trend and local fluctuations of the layout density. At the same time, the graphic density distribution map is trained by a machine learning algorithm (such as a regression model, a classification tree or a neural network), and a mapping relationship between the graphic density and the long-range deviation can be established. This mapping relationship can capture the global deviation trend caused by the uneven distribution of the layout density, and provide data support for the construction of the long-range deviation assessment model.

[0045] After obtaining sufficient statistical features and machine learning models, the long-range deviation evaluation model is further refined by combining physical models and simulation technology. Physical models can include light intensity distribution models, diffraction effect models, and material etching models in the lithography process, which can simulate the physical effects caused by changes in layout density during lithography and etching. By combining these physical models with the output of machine learning algorithms, a long-range deviation evaluation model that considers both physical mechanisms and statistical laws can be constructed. The long-range deviation evaluation model aims to capture and quantify the deviation trends of mask graphics at different spatial scales, especially those deviations that are closely related to changes in graphic density.

[0046] S300: Determine a short-range deviation assessment model according to the graphic information.

[0047] Alternatively, short-range deviation focuses on deviations at the nanometer (nm) scale. This deviation depends not only on the type of pattern (such as shape, size, etc.), but also on the positional relationship between the patterns. Even if two patterns have the same type and size, they may show different deviations if they are located in different graphic environments (such as the density, shape, spacing, etc. of the surrounding graphics). This is because changes in the local environment affect the distribution and reaction rate of the reaction gas and reaction products on the pattern surface, resulting in differences in etching results.

[0048] Optionally, in a feasible implementation of the present application, key features in the graphic information are first extracted, including but not limited to the geometric shape, size parameters, and relative positions and angles of the graphics. This information provides a basis for understanding how the pattern affects the local etching process.

[0049] Next, the graphic information is classified and statistically analyzed to identify the deviation characteristics of different pattern types in the short range. This includes comparing the etching results of the same type of pattern under different local environments, and analyzing the mutual influence of different types of patterns under the same local environment. Through these analyses, a correlation model between pattern type, local environment and short-range deviation can be established.

[0050] In order to build a short-range deviation assessment model, machine learning algorithms or statistical modeling techniques can also be used. These techniques can handle complex nonlinear relationships and extract useful patterns and laws from large amounts of data. During the training process, graphical information in historical data sets can be used as input, and the parameters of the model can be continuously adjusted through iterative optimization algorithms to improve the model's prediction accuracy for short-range deviations.

[0051] Ultimately, the short-range deviation assessment model will be able to predict the deviation of the mask pattern to be corrected within a short range based on the given pattern information. This model not only considers the direct impact of the pattern type on the deviation, but also considers the regulatory effect of the local environment on the deviation. Therefore, it can more accurately reflect the local deviation problems that the mask pattern may encounter during the manufacturing process, and provide strong support for subsequent deviation correction and optimization.

[0052] S400, determining a deviation value of a mask pattern to be corrected according to the long-range deviation evaluation model and the short-range deviation evaluation model.

[0053] Optionally, in a feasible implementation of the present application, the mask pattern to be corrected can first be input into the long-range and short-range deviation evaluation models respectively. The long-range deviation evaluation model focuses on evaluating the impact of changes in pattern density on etching deviations in a global range. It uses the information provided by the pattern density distribution map to predict the overall size deviation caused by uneven density. On the other hand, the short-range deviation evaluation model pays more attention to the local interaction between patterns and the direct impact of pattern types on deviations. The short-range deviation evaluation model uses the positional relationship and pattern type information in the graphic information to predict the specific deviation of the mask pattern to be corrected within a short range.

[0054] The final deviation value not only reflects the global dimensional deviation trend, but also covers the deviation details in the local area due to pattern type and local environmental differences. These deviation values ​​will serve as an important basis for subsequent mask pattern correction and optimization, helping manufacturers improve the manufacturing accuracy and consistency of mask patterns. By comprehensively applying long-range and short-range deviation evaluation models, the comprehensiveness and accuracy of deviation value determination are ensured, providing strong technical support for high-precision fields such as semiconductor manufacturing.

[0055] In a method for determining a mask graphic deviation value provided in an embodiment of the present application, firstly, the detailed graphic information of each graphic in the historical design layout is obtained, which includes but is not limited to the precise positional relationship between the graphics and the corresponding graphic types. This information constitutes the basis of the short-range deviation evaluation model, which can capture the subtle deviations caused by the interaction between local patterns and make up for the shortcomings of the rule table in covering specific patterns. Furthermore, by parsing the graphic density distribution map of the historical design layout, the method constructs a long-range deviation evaluation model, which can evaluate the global deviation trend caused by the overall layout density difference, which is an important aspect ignored by the traditional rule table-based method. The combination of long-range and short-range evaluation models not only focuses on the mutual influence between a single or a few graphics, but also examines the potential contribution of the entire design layout to the mask graphic deviation from a macro perspective, thereby realizing a multi-level and high-precision prediction of the deviation value, thereby improving the accuracy of the determination of the mask graphic deviation value.

[0056] In one embodiment, if Figure 2 As shown, the above step 400 can specifically perform the following steps:

[0057] S410, fusing the long-range deviation assessment model and the short-range deviation assessment model to obtain a target deviation assessment model;

[0058] S420, determining the deviation value of the mask pattern to be corrected according to the target deviation evaluation model.

[0059] Optionally, in a feasible implementation of the present application, the long-range deviation assessment model is fused with the short-range deviation assessment model to construct a more comprehensive and accurate target deviation assessment model, and the deviation value is determined accordingly. Specifically, the fusion process first includes the weight assignment of the output results of the two models, which is based on their respective relative accuracy and importance in predicting the deviation. The weight assignment can be determined by statistical methods, machine learning algorithms, or expert experience to ensure that the advantages of the two models are fully considered in the fusion process.

[0060] Once the weights are determined, the global bias trend predicted by the long-range bias assessment model and the local bias details predicted by the short-range bias assessment model can be weighted summed according to these weights or other appropriate fusion strategies can be adopted. This fusion strategy aims to integrate the prediction capabilities of the two models at different spatial scales, thereby generating a target bias assessment model that reflects both global bias and local details.

[0061] Next, the target deviation evaluation model is used to process the input data of the mask pattern to be corrected. Based on these input data, the target deviation evaluation model combines its internal fusion of long-range and short-range deviation prediction mechanisms to calculate the deviation values ​​of the mask pattern to be corrected at each key position.

[0062] These deviation values ​​not only reflect the global size deviation caused by changes in pattern density, but also capture the local deviation caused by local interactions between patterns and differences in pattern types. Therefore, they can provide more detailed and accurate guidance for subsequent mask pattern correction and optimization.

[0063] In these optional embodiments, the long-range and short-range deviation assessment models are integrated to obtain the target deviation assessment model, which can comprehensively and accurately capture the deviation characteristics of the mask pattern to be corrected at different scales. This method not only improves the comprehensiveness of the deviation assessment, but also enhances the accuracy of the assessment, which helps to correct the mask pattern more accurately in the future, thereby improving the quality and efficiency of the lithography process and reducing the defective product rate in production. It is a key step in optimizing the semiconductor manufacturing process and improving chip performance and reliability.

[0064] In one embodiment, if Figure 3 As shown, the above step 420 can specifically perform the following steps:

[0065] S421, compensating the initial mask pattern corresponding to the design layout to be corrected according to the preset deviation table to obtain a first mask pattern, and determining the first mask pattern as the mask pattern to be corrected, wherein the layout size range of the design layout to be corrected is the same as the layout size range of the historical design layout;

[0066] S422, inputting the design layout to be corrected into the target deviation evaluation model to obtain a target deviation map corresponding to different size ranges of the design layout to be corrected, wherein the target deviation map includes deviation values ​​of the first mask pattern at different positions;

[0067] S423, compensating the first mask pattern according to the target deviation map to obtain a compensated mask pattern corresponding to the design layout to be corrected.

[0068] Optionally, in an embodiment of the present application, the preset deviation table is a table or database containing deviation compensation information, which records the type, size and distribution of deviations that may occur in the mask pattern under different design or manufacturing conditions based on a large amount of historical data and accumulated experience. This table or database is used to guide how to adjust the initial mask pattern to compensate for these potential deviations. In practical applications, the preset deviation table can be continuously updated and optimized to adapt to new designs or changes in manufacturing processes.

[0069] Optionally, in a feasible implementation of the present application, first, according to a preset deviation table, for the design layout to be corrected that has the same size range as the historical design layout, the initial mask pattern corresponding to it is preliminarily compensated. This compensation process is intended to predict and reduce possible systematic deviations based on historical data, thereby generating a first mask pattern that is closer to the ideal state, which is temporarily determined as the starting point of the mask pattern to be corrected.

[0070] Subsequently, the design layout to be corrected is input into the target deviation evaluation model that has been constructed. Through model calculation, a target deviation map is generated, which shows in detail the specific deviation values ​​of the first mask pattern at different positions. This deviation map not only contains the global size deviation trend, but also accurately captures the deviation details caused by the interaction between patterns and the difference in pattern types in the local area, providing a reliable basis for subsequent precise compensation.

[0071] Finally, the first mask pattern is further accurately compensated according to the target deviation map. In this step, the size or shape parameters of the corresponding positions of the mask pattern can be adjusted according to the deviation values ​​of each position in the deviation map to eliminate or reduce these deviations. Through this compensation process, the compensated mask pattern corresponding to the design layout to be corrected is finally generated. This compensated mask pattern not only reduces the global deviation, but also makes fine adjustments to the local area, thereby greatly improving the manufacturing accuracy and consistency of the mask pattern. The entire process, from initial compensation to target deviation assessment, and then to precise compensation, forms a closed-loop deviation correction system, which provides strong technical support and guarantee for high-precision fields such as semiconductor manufacturing.

[0072] In these optional embodiments, the initial mask pattern is preliminarily compensated by using a preset deviation table, and then the target deviation evaluation model is used to accurately predict the deviation and generate a target deviation map. Finally, precise compensation is performed based on the deviation map, which effectively improves the manufacturing accuracy and consistency of the mask pattern, reduces the accumulation of deviations in the manufacturing process, and improves the yield and production efficiency in high-precision manufacturing fields such as semiconductors.

[0073] In one embodiment, the step of obtaining a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout includes:

[0074] According to the historical design map, density matrices of different area sizes are constructed;

[0075] Generate a graphic density distribution map of different area ranges according to the preset generation time and the density matrix of the different area range sizes;

[0076] According to the graphic matching technology, the positional relationship between the graphics in the historical design layout and the graphic type corresponding to each graphic are determined.

[0077] Optionally, in a feasible implementation of the present application, first, key graphic data is extracted from the historical design layout, and these data are used to construct density matrices of multiple different regional sizes. The construction of these density matrices is based on the principles of spatial statistics. By dividing different regional ranges and calculating the distribution density of graphics in each region, a quantitative description of the density distribution of the entire design layout is achieved. The density matrices of different regional ranges can reflect the density change characteristics of different spatial scales in the design layout, providing a rich data basis for subsequent analysis.

[0078] Then, according to the preset generation time, the density matrices of these different area sizes are used to dynamically generate the corresponding graphic density distribution map. By adjusting the generation time parameters, the generation speed and resolution of the density distribution map can be flexibly controlled to meet different analysis requirements. In the process of generating the graphic density distribution map, it is necessary to use optimization algorithms to reduce the calculation time and ensure that the entire process can be completed within a reasonable time range.

[0079] Finally, in order to gain a deeper understanding of the positional relationship and graphic type between the graphics in the historical design layout, graphic matching technology can also be used. Graphic matching technology is a method based on the principles of image processing and pattern recognition. It can determine the relative position and type information between different graphics by comparing the similarities and differences between them. In this application, one or more graphic templates will first be defined, which represent the various types of graphics that may appear in the design layout. Then, these templates are matched with the actual graphics in the design layout, and the specific type of each graphic is identified by calculating the similarity between them. At the same time, a positional relationship network between graphics is constructed based on the relative position information between graphics, which is of great significance for the subsequent analysis of the interaction and deviation causes between graphics.

[0080] In these optional embodiments, by finely analyzing the historical design layout, a multi-scale density matrix is ​​first constructed to fully capture the graphic distribution characteristics, and then a graphic density distribution map is generated by time to facilitate tracking of design evolution and potential problem areas. At the same time, the positional relationship and type between graphics are accurately identified using graphic matching technology, providing a detailed and accurate graphic information foundation for design optimization, error detection and performance evaluation. This significantly improves the depth and efficiency of design analysis, helps to quickly locate design bottlenecks, and promotes the continuous improvement of design quality.

[0081] In one embodiment, after acquiring the graphic density distribution map of the historical design layout and the graphic information of each graphic in the historical design layout, and before determining the long-range deviation assessment model according to the graphic density distribution map, the method further includes:

[0082] Evaluate the full-image key dimension uniformity data of the second mask pattern according to a plurality of key measurement positions, wherein the plurality of key measurement positions are positions in the historical design layout where the variation of the pattern distribution density is greater than a preset threshold, wherein the variation of the pattern distribution density is determined by comparing the pattern distribution densities of adjacent positions, and the full-image key dimension uniformity data is used to evaluate the uniformity of the key dimension of the second mask pattern, wherein the second mask pattern is obtained by compensating the initial mask pattern corresponding to the historical design layout according to a preset deviation table;

[0083] Determining a long-range deviation assessment model according to the graphic density distribution map includes:

[0084] Determining a long-range deviation assessment model according to the graphic density distribution map and the full-image key dimension uniformity data;

[0085] Determining a short-range deviation assessment model according to the graphic information includes:

[0086] A short-range deviation assessment model is determined based on the graphic information and the full-image key dimension uniformity data.

[0087] Optionally, in the embodiment of the present application, the key measurement positions refer to those specific areas or points in the historical design layout where the variation of the pattern density is greater than a preset threshold. These positions are considered "key points" because they are often more susceptible to the impact of the manufacturing process, resulting in significant changes in the size or shape of the mask pattern at these positions. By performing accurate dimensional measurements at these positions, the uniformity of the entire mask pattern can be more effectively evaluated and potential sources of deviation can be identified.

[0088] Optionally, the second mask pattern is a mask pattern obtained by compensating the initial mask pattern corresponding to the historical design layout according to a preset deviation table. The preset deviation table contains deviation information based on historical data and accumulated experience, and is used to guide how to adjust the initial mask pattern to compensate for deviations that may be introduced in the design or manufacturing process. By applying the preset deviation table, it is expected that a mask pattern that is closer to the ideal state, that is, the second mask pattern, can be obtained.

[0089] Optionally, the full-image key dimension uniformity data refers to a quantitative representation of the uniformity of key dimensions between key measurement positions on the entire second mask pattern. These key dimensions include line width, spacing, side length of the pattern, etc., which are important indicators for evaluating the dimensional accuracy of the mask pattern. The full-image key dimension uniformity data is obtained by statistically analyzing the key dimension data at these key measurement positions, and is used to evaluate the dimensional stability of the first mask pattern in the global range.

[0090] It should be noted that although the preset deviation table is based on a large amount of historical data and accumulated experience, it may still have certain limitations in practical applications. This is because each design layout has its own unique characteristics and complexity, and the preset deviation table may not be able to fully cover all possible situations. Therefore, in practical applications, the use of the preset deviation table for preliminary compensation is mainly to obtain a relatively reasonable initial mask pattern. Although this second mask pattern may still have certain deviations, it is already closer to the ideal state than the uncompensated initial mask pattern, providing a better starting point for subsequent more precise adjustments and corrections.

[0091] Optionally, in a feasible implementation of the present application, the areas where the variation of the graphic density exceeds the preset threshold can be automatically identified based on the graphic density distribution map of the historical design layout, and these areas are marked as key measurement positions. Subsequently, high-precision measurement tools or algorithms can be used to accurately measure the key dimensions of these key measurement positions. These key dimensions include line width, spacing, side length of the graphic, etc., which are important indicators for evaluating the accuracy of the mask graphic dimensions.

[0092] During the measurement process, the key dimension data of each key measurement position is recorded and statistical analysis is performed to evaluate the size uniformity of the second mask pattern at these positions. The second mask pattern here is obtained after the preset deviation compensation is performed on the initial mask pattern corresponding to the historical design layout. The purpose of compensation is to minimize the deviation caused by the design or manufacturing process. By evaluating the key dimension uniformity of the compensated mask pattern, the compensation effect can be further verified and the deviation areas that may still exist can be identified.

[0093] When evaluating the uniformity of key dimensions of the entire image, the size differences between different key measurement locations can be considered, and statistics such as standard deviation and coefficient of variation can be calculated to quantify the size uniformity of the mask pattern in the global range. These data not only reflect the dimensional stability of the mask pattern in different areas, but also provide an important basis for the subsequent construction of long-range and short-range deviation evaluation models.

[0094] It should be noted that in determining the uniformity data of the key dimension of the entire second mask pattern, it is not necessary to directly measure every point or crystal on the second mask pattern, because that would be an extremely large and unrealistic task. Instead, the present application measures by selecting key measurement positions, which are determined based on the variation of the pattern density greater than a preset threshold, and they represent areas in the design layout that may be greatly affected by the manufacturing process.

[0095] By accurately measuring the dimensions of these key measurement positions, critical dimension data about these positions can be obtained. Then, using these data, the critical dimension uniformity of the second mask pattern in the entire range can be further evaluated. This evaluation can specifically involve statistical analysis methods, such as calculating standard deviations, coefficients of variation, etc., to quantitatively represent the dimensional stability of the mask pattern in the global range. The critical dimension uniformity of the entire second mask pattern is effectively evaluated through a limited number of key measurement positions, thereby providing reliable data support for subsequent deviation evaluation and correction. This method not only improves efficiency, but also ensures the accuracy and representativeness of the evaluation results.

[0096] Optionally, in a feasible implementation of the present application, the pattern density distribution diagrams of different regions are first analyzed. By comparing the density differences in different regions, local or global density change patterns that may exist in the design layout can be identified. These patterns are often associated with deviation sources in the manufacturing process, such as uneven exposure and etching rate differences during the photolithography process.

[0097] Next, the analysis results of the pattern density distribution map can be combined with the full-pattern key dimension uniformity data of the second mask pattern. By correlating these two data sets, the potential relationship between pattern density variation and key dimension deviation can be identified.

[0098] Based on this relationship, a long-range bias assessment model can be constructed. This model aims to capture and quantify the bias trends of mask patterns at different spatial scales, especially those that are closely related to pattern density changes.

[0099] In the process of building a long-range deviation assessment model, other possible influencing factors can also be considered, such as manufacturing process parameters, material properties, etc. These factors may interact with pattern density and key dimension deviation, thereby affecting the prediction performance of the model. Therefore, a variety of factors can be considered comprehensively, and a long-range deviation assessment model that is both accurate and practical can be constructed through multivariate analysis and optimization algorithms.

[0100] Ultimately, the long-range deviation assessment model will be able to predict the deviation of the mask pattern to be corrected at different spatial scales, providing an important reference for subsequent deviation correction and optimization. The realization of this step not only improves the accuracy and stability of mask pattern manufacturing, but also lays a solid foundation for improving the performance and quality of the final product.

[0101] Optionally, in a feasible implementation method of the present application, the graphic information extracted from the historical design layout, including the type, positional relationship, shape characteristics, etc. of the graphics, is first used as the basic data for model input. At the same time, the graphic information and the key dimension uniformity data of the whole image are input into the model in combination with the key dimension uniformity data of the whole image obtained by evaluating multiple key measurement positions. By analyzing and learning these data, the model can capture the influence of local interactions between graphics and changes in graphic density on short-range deviations. During the training process, the model parameters are continuously adjusted and the model structure is optimized to improve the accuracy of its prediction of short-range deviations. Finally, after training and optimization, a short-range deviation assessment model that can comprehensively consider graphic information and key dimension uniformity data of the whole image is obtained.

[0102] Alternatively, if Figure 4As shown in the figure, it can be seen that in the long-term loading mode, the critical dimension after etching (AEI) is mainly related to the change of pattern density. In the short-term loading mode, the critical dimension after etching mainly depends on the type of pattern (such as shape, size, etc.) and the positional relationship between each pattern.

[0103] In these optional embodiments, by comprehensively considering the graphic density distribution of the historical design layout, graphic information, and the full-image key dimension uniformity data of the key measurement positions, the accuracy and practicality of the long-range and short-range deviation evaluation models are significantly improved.

[0104] In one embodiment, determining the long-range deviation assessment model according to the graphic density distribution map and the full-image key dimension uniformity data includes:

[0105] Determine an actual deviation map of the second mask pattern according to the post-development size of the second mask pattern and the full-pattern key dimension uniformity data, wherein the actual deviation map is used to indicate the deviation value between the size of each position of the second mask pattern and the preset size;

[0106] The first basic deviation assessment model to be trained is trained according to the actual deviation map and the graphic density distribution map to obtain the long-range deviation assessment model, and the long-range deviation assessment model is used to generate a deviation map corresponding to the graphic density distribution map of the first size range according to the graphic density distribution map of the first size range.

[0107] Optionally, although graphic density distribution maps of different area range sizes are generated before, these area ranges range from very local small areas to large areas of the entire design layout. And each area range contains graphic elements of different size ranges. In other words, the graphic density distribution maps of different area range sizes are actually the results of sampling and analyzing graphic elements under different size ranges. Therefore, these graphic density distribution maps actually contain graphic density information of different size ranges. When training the long-range deviation evaluation model, you can select the appropriate area range and size range as needed to more accurately evaluate the long-range deviation.

[0108] In these optional embodiments, by combining the pattern density distribution map with the full-image key dimension uniformity data, an actual deviation map is accurately constructed to reflect the size deviation of each position of the mask pattern. Subsequently, these actual data and the pattern density distribution map are used to train the basic deviation evaluation model, thereby optimizing the long-range deviation evaluation model. This model can efficiently predict the deviation map for the pattern density distribution map in a specific size range, improve the accuracy of deviation prediction in the lithography process and the ability to control long-range deviations, which is crucial to ensuring the quality of semiconductor manufacturing.

[0109] In one embodiment, determining a short-range deviation assessment model according to the graphic information and the full-image key dimension uniformity data includes:

[0110] The second basic deviation assessment model to be trained is trained according to the actual deviation map and the graphic information to obtain a trained short-range deviation assessment model, and the short-range deviation assessment model is used to generate a deviation map corresponding to the graphic information in a second size range according to the graphic information in the second size range, and the second size range is smaller than the first size range.

[0111] Optionally, in a feasible implementation of the present application, first, for the construction of the long-range deviation evaluation model, starting from the post-development size data of the second mask pattern, combined with the key size uniformity data of the whole pattern, the actual deviation map of the second mask pattern is calculated. This actual deviation map is a key intermediate product, which records in detail the actual deviation value of each position on the mask pattern relative to the preset size, providing accurate label data for subsequent model training.

[0112] Next, the first basic deviation assessment model is trained using these actual deviation maps and the graphic density distribution map previously constructed based on the historical design layout as training input. The graphic density distribution map, as a feature input, reflects the density of the distribution of the graphics in space, which is crucial for understanding the size deviation caused by uneven light intensity distribution during the lithography process. Through repeated iterative training, the model parameters are continuously optimized until the model can accurately predict the deviation map corresponding to the graphics within any given first size range based on the graphic density distribution map of the first size range, thereby obtaining a trained long-range deviation assessment model. This process effectively establishes a long-range correlation between graphic density and lithography deviation, providing strong support for the prediction of large-size deviations in the lithography process.

[0113] A similar strategy is used to build the short-range deviation assessment model, but it focuses more on the local interactions and influences between graphics. It is also based on the actual deviation map, but this time the graphic information is used as training input to train the second basic deviation assessment model. Graphic information provides a more detailed description of local features than the graphic density distribution map, which helps the model capture the short-range impact of small changes between graphics on lithography deviations. Through training, a short-range deviation assessment model is obtained that can generate corresponding deviation maps for graphic information in the second size range (smaller than the first size range). This model has important application value in predicting local lithography deviations, optimizing graphic layout, and reducing mutual influence between graphics.

[0114] For example, the first size range can be set to micrometer (μm) to millimeter (mm) level, which is used to describe larger-scale graphic structures on the chip, such as macro layout planning or larger functional areas. At this scale, the deviation in the photolithography process is often affected by multiple factors such as pattern density, light intensity distribution, mask error, etc., and these factors show more significant interactions in the long range.

[0115] The second size range is set to the nanometer (nm) level, which is a core concern in the semiconductor manufacturing process because the feature size of modern integrated circuits has been reduced to tens of nanometers or even a few nanometers. At the nanoscale, the prediction and control of lithography deviations become more complex and critical because it directly affects the performance, power consumption and reliability of the chip. Slight changes between graphics, such as edge roughness, shape distortion, etc., may have a significant impact on the final product quality. It should be noted that the above-mentioned setting of the first size range and the second size range is only an exemplary description, which aims to illustrate the possible application scenarios of the long-range deviation evaluation model and the short-range deviation evaluation model in semiconductor manufacturing. In fact, the selection of these two size ranges is not fixed, but can be flexibly adjusted according to the specific manufacturing process requirements, chip design complexity and the capabilities of lithography technology. The key is that no matter which size range is selected, it should be ensured that the first size range is larger than the second size range.

[0116] In these optional embodiments, by combining the actual deviation map with the detailed graphic information, a short-range deviation assessment model is trained, which can accurately predict the corresponding deviation map for graphic information in a small size range such as nanometer level.

[0117] Optionally, in a complete embodiment of the present application, the present application proposes a loading proximity correction (LPC) scheme, which aims to comprehensively and accurately reflect the loading status at the full chip level, especially for pattern loading effects that are difficult to cover with prior art and are not placed in the rule table.

[0118] Specifically, this application is implemented in two major stages: First, in the "Long Term Loading" stage, by constructing a mapping relationship between graphic density and loading deviation, and introducing an artificial intelligence algorithm to automatically search for a variety of graphic densities, a flexible DAL (Density+AI+Loading) model (i.e., a long-range deviation evaluation model) is established. This long-range deviation evaluation model can effectively capture the impact of different density distributions on the transport of reactive gases and products during the etching process, and then explain the resulting inconsistent etching rates and loading deviations.

[0119] Next, in the "Local Term Loading" stage, the influence of pattern type and its environment on loading deviation was further considered, and the PDAL (Pattern Based+AI+Loading) model (i.e., short-term deviation assessment model) was established. This short-term deviation assessment model uses AI algorithms to automatically explore pattern environments of different ranges and types, making the model more flexible and able to accurately reflect local loading effects. Local loading is not only related to the type characteristics of the pattern, such as shape and size, but is also deeply affected by the surrounding pattern environment. Even patterns of the same type will show different loading deviations in different environments (i.e., the positional relationship between the graphics).

[0120] In specific implementation, the LPC scheme first uses the Bias Table (i.e., the preset deviation table) to obtain the initial ADI target value (i.e., the size after development), and then identifies the pattern area that contributes more to loading (i.e., the key measurement position) through GDS sampling for key measurement. A sampling tool is generated based on pattern matching technology to collect critical dimension uniformity (CDU) mapping and graphic information of these areas, and the density map calculation time is shortened to an acceptable range through an optimization algorithm. Next, the density matrix of multiple ranges and grids is constructed and optimized to enhance the flexibility and adaptability of the model. Subsequently, the DAL (i.e., long-range deviation evaluation model) and PDAL model (i.e., short-range deviation evaluation model) are established respectively. The former focuses on the relationship between global density and loading deviation, while the latter focuses on the influence of local pattern type and environment on loading deviation.

[0121] Finally, the PDAL model is integrated into the DAL model to achieve coupled optimization of the two. By obtaining the deviation map and density map from the two models, retargeting is performed in segments, and the deviation map is applied to the lithography image, thereby achieving accurate correction of the loading state at the full chip level. This comprehensive solution not only improves the accuracy and stability of the lithography process, but also provides strong technical support for deviation control in semiconductor manufacturing, which helps to improve the overall performance and yield of the chip.

[0122] In these optional embodiments, the LPC solution proposed in one embodiment of the present application achieves comprehensive coverage and high-precision correction of chip loading deviations through the DAL and PDAL models, significantly surpassing the traditional SSA table technical solution. First, in terms of coverage, the DAL model covers long-term loading at the micron level and even the millimeter level into its compensation scope. The key to this capability is that the DAL model supports flexible adjustment of the size range and can accurately capture the impact of changes in image density within different scale ranges on loading deviations. In fact, the loading deviations at the micron and millimeter levels differ significantly, and existing SSA table technology often cannot effectively identify these differences, resulting in limited compensation effects. In addition, for short-range deviations at the nanometer level, the PDAL model effectively solves the problem of differences in loading deviations of the same graphic in different environments with its sophisticated fragment-level analysis, which is beyond the reach of traditional SSA table technology.

[0123] In terms of accuracy, thanks to comprehensive coverage and sophisticated modeling technology, it can accurately reflect the loading deviation at the full chip level. In actual verification, the model error (the difference between the predicted deviation value and the actual measured deviation value) of the above-mentioned long-range deviation evaluation model reached 97% of the sample errors less than 0.25nm, and 99% of the sample errors less than 0.5nm.

[0124] Finally, the generalization ability of LPC technology is also a highlight. As a model-based technical solution, LPC cleverly incorporates AI algorithms, enabling it to flexibly respond to various complex situations. The combination of DAL and PDAL models not only establishes a massive mapping relationship between graphic density, type, environment and loading deviation, but also enables LPC to predict loading deviations in similar size ranges under the same technology node. Users only need to input the GDS file of the design layout to quickly obtain accurate full-chip-level loading deviation values, greatly saving time, manpower and costs. In contrast, traditional SSA table technology requires frequent collection of test graphics and manual value supplementation, which is a cumbersome process and has limited accuracy.

[0125] In summary, the LPC technology solution has brought revolutionary changes to the semiconductor manufacturing industry with its comprehensive coverage, high-precision compensation capabilities and strong generalization capabilities, and has provided strong technical support for improving product quality and production efficiency.

[0126] Figure 5 A schematic structural diagram of a mask pattern deviation value determination device provided in another embodiment of the present application is shown. For ease of explanation, only the portion related to the embodiment of the present application is shown.

[0127] Reference Figure 5 , the mask pattern deviation value determining device may include:

[0128] An acquisition module 501 is used to acquire a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout, wherein the graphic information includes a positional relationship between each graphic and a graphic type corresponding to each graphic;

[0129] A first determination module 502 is used to determine a long-range deviation assessment model according to the graphic density distribution map;

[0130] A second determination module 503 is used to determine a short-range deviation assessment model according to the graphic information;

[0131] The third determination module 504 is used to determine the deviation value of the mask pattern to be corrected according to the long-range deviation evaluation model and the short-range deviation evaluation model.

[0132] In one embodiment, the third determining module 504 may include:

[0133] A fusion submodule, used to fuse the long-range deviation assessment model and the short-range deviation assessment model to obtain a target deviation assessment model;

[0134] The first determination submodule is used to determine the deviation value of the mask pattern to be corrected according to the target deviation evaluation model.

[0135] In one embodiment, the first determining submodule may include:

[0136] A first compensation unit is used to compensate the initial mask pattern corresponding to the design layout to be corrected according to a preset deviation table to obtain a first mask pattern, and determine the first mask pattern as the mask pattern to be corrected, wherein the layout size range of the design layout to be corrected is the same as the layout size range of the historical design layout;

[0137] A first determination unit is used to input the to-be-corrected design layout into the target deviation evaluation model to obtain a target deviation map corresponding to different size ranges of the to-be-corrected design layout, wherein the target deviation map includes deviation values ​​of the first mask pattern at different positions;

[0138] The second compensation unit is used to compensate the first mask pattern according to the target deviation map to obtain a compensated mask pattern corresponding to the design layout to be corrected.

[0139] In one embodiment, the acquisition module 501 may include:

[0140] A construction submodule, used for constructing density matrices of different area sizes according to the historical design layout;

[0141] A generating submodule, for generating a graphic density distribution map of different area ranges according to a preset generating time and the density matrix of the different area range sizes;

[0142] The second determination submodule is used to determine the positional relationship between the graphics in the historical design layout and the graphic type corresponding to each graphic according to the graphic matching technology.

[0143] In one embodiment, the mask pattern deviation value determining device may further include:

[0144] A fourth determination module is used to evaluate the full-image key dimension uniformity data of the second mask graphic according to multiple key measurement positions, wherein the multiple key measurement positions are positions in the historical design layout where the variation of the graphic distribution density is greater than a preset threshold, and the variation of the graphic distribution density is determined by comparing the graphic distribution densities of adjacent positions. The full-image key dimension uniformity data is used to evaluate the uniformity of the key dimension of the second mask graphic, and the second mask graphic is obtained by compensating the initial mask graphic corresponding to the historical design layout according to a preset deviation table.

[0145] In one embodiment, the first determining module 502 may include:

[0146] A third determination submodule is used to determine a long-range deviation evaluation model according to the graphic density distribution map and the full-image key dimension uniformity data;

[0147] In one embodiment, the second determining module 503 may include:

[0148] The fourth determination submodule is used to determine a short-range deviation evaluation model according to the graphic information and the full-image key dimension uniformity data.

[0149] In one embodiment, the third determining submodule may include:

[0150] A second determining unit, configured to determine an actual deviation map of the second mask pattern according to the post-development size of the second mask pattern and the full-image key dimension uniformity data, wherein the actual deviation map is used to indicate a deviation value between a size of each position of the second mask pattern and a preset size;

[0151] The first training unit is used to train the first basic deviation assessment model to be trained according to the actual deviation map and the graphic density distribution map to obtain the long-range deviation assessment model, and the long-range deviation assessment model is used to generate a deviation map corresponding to the graphic density distribution map of the first size range according to the graphic density distribution map of the first size range.

[0152] In one embodiment, the fourth determining submodule may include:

[0153] The second training unit is used to train a second basic deviation assessment model to be trained according to the actual deviation map and the graphic information to obtain a trained short-range deviation assessment model, and the short-range deviation assessment model is used to generate a deviation map corresponding to the graphic information in a second size range according to the graphic information in the second size range, and the second size range is smaller than the first size range.

[0154] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application, and are devices corresponding to the above-mentioned methods. All implementation methods in the above-mentioned method embodiment are applicable to the embodiments of the device. Its specific functions and technical effects brought about can be found in the method embodiment part, and will not be repeated here.

[0155] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0156] Figure 6 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0157] The device may include a processor 601 and a memory 602 storing program instructions.

[0158] When the processor 601 executes the program, the steps in any of the above method embodiments are implemented.

[0159] Exemplarily, the program may be divided into one or more modules / units, one or more modules / units are stored in the memory 602 and executed by the processor 601 to complete the present application. One or more modules / units may be a series of program instruction segments capable of completing a specific function, and the instruction segments are used to describe the execution process of the program in the device.

[0160] Specifically, the processor 601 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0161] The memory 602 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 602 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 602 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 602 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 602 is a non-volatile solid-state memory.

[0162] The memory may include read-only memory (ROM), random access memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical or other physical / tangible memory storage devices. Thus, typically, the memory includes one or more tangible (non-transitory) readable storage media (e.g., memory devices) encoded with software including computer executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0163] The processor 601 implements any one of the methods in the above embodiments by reading and executing program instructions stored in the memory 602 .

[0164] In one example, the electronic device may further include a communication interface 603 and a bus 610. The processor 601, the memory 602, and the communication interface 603 are connected via the bus 610 and communicate with each other.

[0165] The communication interface 603 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0166] Bus 610 includes hardware, software or both, and the parts of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 610 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0167] In addition, in combination with the method in the above embodiment, the embodiment of the present application can provide a storage medium for implementation. The storage medium stores program instructions; when the program instructions are executed by a processor, any one of the methods in the above embodiment is implemented.

[0168] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0169] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0170] An embodiment of the present application provides a computer program product, which is stored in a storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0171] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0172] The functional modules shown in the above block diagram can be implemented as hardware, software, firmware or their combination. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), suitable firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. Programs or code segments can be stored in machine-readable media, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable media" can include any medium that can store or transmit information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. Code segments can be downloaded via computer grids such as the Internet, intranets, etc.

[0173] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0174] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0175] The above are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A method for determining a mask pattern deviation value, characterized in that: The method comprises: Obtaining a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout, wherein the graphic information includes a positional relationship between each graphic and a graphic type corresponding to each graphic; Determining a long-range deviation assessment model based on the graphic density distribution map; Determining a short-range deviation assessment model based on the graphical information; The deviation value of the mask pattern to be corrected is determined according to the long-range deviation evaluation model and the short-range deviation evaluation model.

2. The method for determining the mask pattern deviation value according to claim 1, characterized in that: The step of determining the deviation value of the mask pattern to be corrected according to the long-range deviation evaluation model and the short-range deviation evaluation model comprises: The long-range deviation assessment model and the short-range deviation assessment model are integrated to obtain a target deviation assessment model; The deviation value of the mask pattern to be corrected is determined according to the target deviation evaluation model.

3. The method for determining the deviation value of a mask pattern according to claim 2, characterized in that: Determining the deviation value of the mask pattern to be corrected according to the target deviation evaluation model includes: According to the preset deviation table, the initial mask pattern corresponding to the design layout to be corrected is compensated to obtain a first mask pattern, and the first mask pattern is determined as the mask pattern to be corrected, wherein the layout size range of the design layout to be corrected is the same as the layout size range of the historical design layout; Inputting the to-be-corrected design layout into the target deviation evaluation model to obtain a target deviation map corresponding to different size ranges of the to-be-corrected design layout, wherein the target deviation map includes deviation values ​​of the first mask pattern at different positions; The first mask pattern is compensated according to the target deviation map to obtain a compensated mask pattern corresponding to the design layout to be corrected.

4. The method for determining a mask pattern deviation value according to claim 1, characterized in that: The step of obtaining a graphic density distribution map of a historical design layout and graphic information of each graphic in the historical design layout includes: According to the historical design map, density matrices of different area sizes are constructed; Generate a graphic density distribution map of different area ranges according to the preset generation time and the density matrix of the different area range sizes; According to the graphic matching technology, the positional relationship between the graphics in the historical design layout and the graphic type corresponding to each graphic are determined.

5. The method for determining the deviation value of a mask pattern according to claim 1, characterized in that: After acquiring the graphic density distribution map of the historical design layout and the graphic information of each graphic in the historical design layout, and before determining the long-range deviation evaluation model according to the graphic density distribution map, the method further includes: Evaluate the full-image key dimension uniformity data of the second mask pattern according to a plurality of key measurement positions, wherein the plurality of key measurement positions are positions in the historical design layout where the variation of the pattern distribution density is greater than a preset threshold, wherein the variation of the pattern distribution density is determined by comparing the pattern distribution densities of adjacent positions, and the full-image key dimension uniformity data is used to evaluate the uniformity of the key dimension of the second mask pattern, wherein the second mask pattern is obtained by compensating the initial mask pattern corresponding to the historical design layout according to a preset deviation table; Determining a long-range deviation assessment model according to the graphic density distribution map includes: Determining a long-range deviation assessment model according to the graphic density distribution map and the full-image key dimension uniformity data; Determining a short-range deviation assessment model according to the graphic information includes: A short-range deviation assessment model is determined based on the graphic information and the full-image key dimension uniformity data.

6. The method for determining the deviation value of a mask pattern according to claim 5, characterized in that: Determining a long-range deviation evaluation model according to the graphic density distribution map and the full-image key dimension uniformity data includes: Determine an actual deviation map of the second mask pattern according to the post-development size of the second mask pattern and the full-pattern key dimension uniformity data, wherein the actual deviation map is used to indicate the deviation value between the size of each position of the second mask pattern and the preset size; The first basic deviation assessment model to be trained is trained according to the actual deviation map and the graphic density distribution map to obtain the long-range deviation assessment model, and the long-range deviation assessment model is used to generate a deviation map corresponding to the graphic density distribution map of the first size range according to the graphic density distribution map of the first size range.

7. The method for determining the deviation value of a mask pattern according to claim 6, characterized in that: Determining a short-range deviation assessment model according to the graphic information and the full-image key dimension uniformity data includes: The second basic deviation assessment model to be trained is trained according to the actual deviation map and the graphic information to obtain a trained short-range deviation assessment model, and the short-range deviation assessment model is used to generate a deviation map corresponding to the graphic information in a second size range according to the graphic information in the second size range, and the second size range is smaller than the first size range.

8. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the method for determining the mask pattern deviation value according to any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method for determining a mask pattern deviation value according to any one of claims 1 to 7 is implemented.

10. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the method for determining a mask pattern deviation value as described in any one of claims 1 to 7.

Images