CN116819911B

CN116819911B - Mask pattern optimization method, mask pattern optimization device, exposure equipment and storage medium

Info

Publication number: CN116819911B
Application number: CN202311111057.4A
Authority: CN
Inventors: 和琨; 牛志元; 陈健; 杜德川
Original assignee: Guangke Xintu Beijing Technology Co ltd
Current assignee: Guangke Xintu Beijing Technology Co ltd
Priority date: 2023-08-31
Filing date: 2023-08-31
Publication date: 2023-10-31
Anticipated expiration: 2043-08-31
Also published as: CN116819911A

Abstract

The application relates to the technical field of photoetching of semiconductor manufacture and discloses a mask pattern optimizing method, a device, exposure equipment and a storage medium.

Description

Mask pattern optimization method, mask pattern optimization device, exposure equipment and storage medium

Technical Field

The present application relates to the field of photolithography technology for semiconductor manufacturing, and in particular, to a mask pattern optimization method, apparatus, exposure device, and storage medium.

Background

The optimization of mask patterns is used as a core technology of computational holographic lithography (also called computational lithography), and the patterns on the mask are optimized mainly by a computer, so that the optimized mask can form a desired design pattern on a wafer after diffraction and interference of light occur under the irradiation of a real light source.

At present, a direct search algorithm with simple operation and good compatibility is adopted in the industry to optimize mask patterns on a common binary mask. Compared with other optimization methods, the binary mask optimized by the direct search algorithm is better in imaging quality and diffraction efficiency. However, the direct search algorithm has a large search space and high computational complexity, resulting in a large amount of time consumed in optimizing the binary mask using the direct search algorithm.

Therefore, how to improve the efficiency of optimization of a mask pattern when the optimization of the mask pattern is performed using a direct search algorithm has become a problem to be solved.

Disclosure of Invention

In view of the above, the present application provides a method, apparatus, exposure apparatus, and storage medium for optimizing a mask pattern, which solve the problem of how to improve the efficiency of optimization of a mask pattern when optimization of a mask pattern is performed using a direct search algorithm.

In a first aspect, the present application provides a method for optimizing a mask pattern, applied to an exposure apparatus of a chip manufacturing process, where the exposure apparatus includes a master node and at least one slave node, each node pre-stores a preset mask pattern and a local pattern in a target pattern, the local patterns stored in the different nodes are different, and the local patterns stored in all the nodes respectively form the target pattern together; the method is performed by a master node, the method comprising:

updating the pixel value of a target pixel point in a preset mask pattern in the current iteration period;

determining the main node similarity between a sub-imaging pattern corresponding to a first local pattern and the first local pattern according to a preset similarity calculation rule, wherein the first local pattern is a local pattern stored by a main node, the sub-imaging pattern is a local imaging pattern in the first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on a pixel value of a target pixel point and a pre-stored preset mask pattern;

distributing the pixel value of the target pixel point to different slave nodes so that the different slave nodes respectively generate a new sub-imaging pattern corresponding to the local pattern stored by the slave nodes according to the pixel value of the target pixel point and a preset mask pattern, and determining the slave node similarity between the local pattern stored by the slave nodes and the new sub-imaging pattern;

Obtaining slave node similarity between self-stored local patterns and new sub-imaging patterns fed back by all slave nodes respectively;

determining an optimization result of a preset mask pattern in a current iteration period according to the similarity of the master node, the similarity of all slave nodes respectively determined by the slave nodes and the initial similarity, wherein the initial similarity is the similarity between a second imaging pattern and a target pattern generated by the preset mask pattern before updating under the target light field;

determining whether the optimization result accords with a preset iterative optimization condition;

when the optimization result is determined to be not in accordance with the preset iterative optimization condition, entering the next iterative period;

or stopping the optimization operation when the optimization result is determined to meet the preset iterative optimization condition.

In the above technical solution, a master node and at least one slave node are set in the exposure apparatus, and each node is set to pre-store a preset mask pattern consistent with other nodes and a local pattern which is different from other nodes and is a part of the target graph, and the local patterns respectively stored in all nodes are guaranteed to form the target pattern together, so that the different nodes in the system can all perform independent operation according to the data stored in the node. In this way, in the current iteration period, the master node updates the pixel value of the target pixel point in the preset mask pattern, and after distributing the pixel value of the target pixel point in the preset mask pattern to different slave nodes, the different slave nodes can independently and parallelly generate a new sub-imaging pattern corresponding to the local pattern stored by themselves according to the pixel value of the target pixel point and the preset mask pattern, and determine the similarity of the local pattern stored by themselves and the new sub-imaging pattern. Meanwhile, the main node also determines the similarity between the sub-imaging pattern corresponding to the first local pattern stored by the main node and the first local pattern according to a preset similarity calculation rule. And obtaining the slave node similarity between the self-stored local patterns fed back by all the slave nodes and the new sub-imaging patterns respectively, so as to determine the optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarity respectively determined by all the slave nodes and the initial similarity. When the optimization result does not accord with the preset iteration optimization condition, entering the next iteration period; or stopping the optimization operation to obtain a final optimization result when the optimization result meets the preset iterative optimization condition. In each iteration, the master node and at least one slave node can simultaneously and parallelly calculate the similarity between the local pattern stored by the master node and the sub-imaging patterns corresponding to the local pattern, so that an optimization result of a preset mask pattern is determined, the traditional method for optimizing the mask pattern based on a direct search algorithm is not needed, the calculation of the similarity between the whole imaging pattern corresponding to the mask pattern and the whole target pattern is shortened to 1/N, N is the number of all nodes in the exposure equipment, and the parallel execution of the direct search algorithm is realized, so that the optimization efficiency of the mask pattern is accelerated.

In some alternative embodiments, before updating the pixel value of the target pixel point in the preset mask pattern in the current iteration period, the method further includes:

dividing the target pattern into a preset number of partial patterns, the preset number being equal to the number of nodes included in the exposure apparatus;

the other partial patterns than the first partial pattern are distributed to different slave nodes, respectively.

In the above technical solution, before updating a preset mask pattern, a master node divides a target pattern into partial patterns with the number equal to the number of nodes included in an exposure device, and distributes other partial patterns except for a first partial pattern to different slave nodes respectively, so as to ensure that each slave node can acquire one partial pattern, and further, when optimizing the mask pattern, all nodes of the exposure device can continuously and parallelly calculate the similarity between the partial pattern stored by the node and the corresponding sub-imaging pattern, thereby saving the time when optimizing the mask pattern by using a direct search algorithm and submitting the optimization efficiency of the mask source.

In some alternative embodiments, updating the pixel value of the target pixel point in the pre-stored preset mask pattern includes:

When the preset mask pattern is the mask pattern in the binary phase mask, selecting a target pixel point in the preset mask pattern to perform phase change operation based on a direct search algorithm so as to modify the pixel value of the target pixel point;

or when the preset mask pattern is the mask pattern in the binary amplitude mask, selecting the target pixel point in the preset mask pattern to carry out amplitude change operation based on a direct search algorithm so as to modify the pixel value of the target pixel point.

In the technical scheme, when the preset mask pattern is the mask pattern in the mask plates of different types, different slave strategies are adopted for updating the preset mask pattern, and the practical application scenes of the mask plates of different types are considered, so that the application scene of mask pattern optimization is widened.

In some optional embodiments, determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to a preset similarity calculation rule includes:

acquiring a pixel value of each pixel point in the first local pattern;

acquiring a pixel value of each pixel point of the sub-imaging pattern corresponding to the first local pattern;

and determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to the pixel values of all the pixel points in the first local pattern and the pixel values of all the pixel points in the sub-imaging pattern corresponding to the first local pattern.

In some optional embodiments, determining an optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes, and the initial similarity includes:

determining the total similarity between the first imaging pattern and the target pattern according to the similarity of the master node and the similarity of all the slave nodes respectively determined by the slave nodes;

and determining an optimization result according to the initial similarity and the total similarity.

In the above technical solution, the master node determines the total similarity between the first imaging pattern and the target pattern according to the master node similarity and the slave node similarities respectively determined by all the slave nodes, and the master node is not required to directly calculate the total similarity between the whole imaging pattern corresponding to the mask pattern and the whole target pattern. Each node can determine a similarity, and then when the total similarity is determined, the calculation efficiency of the total similarity is enabled to be hooked with the number of the master nodes and the slave nodes, parallel calculation of the total similarity is achieved, and when the direct search algorithm is used for optimizing the mask pattern, parallelization design of the direct search algorithm is achieved, and therefore optimization efficiency of the mask pattern is accelerated.

In some alternative embodiments, determining the optimization result from the initial similarity and the overall similarity includes:

when the initial similarity is greater than or equal to the total similarity, updating the value of the initial similarity to the value of the total similarity, and determining the updated preset mask pattern as an optimization result.

In the above technical solution, when the initial similarity is greater than or equal to the total similarity, the preset mask pattern corresponding to the imaging pattern closer to the target pattern is used as the optimization result, and the initial similarity is updated to facilitate the determination of the optimization result in the next iteration period, so as to optimize the mask pattern. And moreover, the optimization result can be determined through simple comparison of the magnitude relation, the optimization result determination flow is simplified, and the optimization efficiency of the mask pattern is further improved when the mask pattern is optimized by using a direct search algorithm.

In some alternative embodiments, when the initial similarity is less than the total similarity, the method further comprises:

restoring the updated preset mask pattern in the current iteration period into the mask pattern acquired in the previous iteration period, taking the mask pattern acquired in the previous iteration period as the original mask pattern of the next iteration period, and entering the optimization operation of the next iteration period to acquire a new iteration optimization result again;

Or when the current iteration period is the first iteration period, the updated preset mask pattern in the current iteration period is restored to the preset initial mask pattern.

In the above technical solution, if the initial similarity is smaller than the total similarity, it means that the current update result is not ideal, at this time, the updated preset mask pattern is restored to the mask pattern obtained in the previous iteration period, and the mask pattern obtained in the previous iteration period is used as the original mask pattern in the next iteration period, and the optimization operation in the next iteration period is performed, so as to obtain a new iteration optimization result again, thereby facilitating the optimization in the next iteration period. The optimization result of the next time or even a plurality of iteration cycles after the next time is not influenced by the non-ideal optimization, and the accuracy of the optimization result is improved.

In a second aspect, the present application provides an optimizing apparatus for mask patterns, applied to an exposure device of a chip manufacturing process, where the exposure device includes a master node and at least one slave node, each node is pre-stored with a preset mask pattern and a local pattern in a target pattern, the local patterns stored in the different nodes are different, and the local patterns stored in all the nodes respectively form the target pattern together; the apparatus is operated on a master node, the apparatus comprising:

The updating module is used for updating the pixel value of the target pixel point in the preset mask pattern in the current iteration period;

the first determining module is used for determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to a preset similarity calculation rule, wherein the first local pattern is a local pattern stored by the main node, the sub-imaging pattern is a local imaging pattern in the first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on the pixel value of the target pixel point and a pre-stored preset mask pattern;

the distribution module is used for distributing the pixel value of the target pixel point to different slave nodes so that the different slave nodes respectively generate a new sub-imaging pattern corresponding to the local pattern stored by the slave nodes according to the pixel value of the target pixel point and a preset mask pattern, and determine the slave node similarity between the local pattern stored by the slave nodes and the new sub-imaging pattern;

the acquisition module is used for acquiring the slave node similarity between the self-stored local patterns and the new sub-imaging patterns fed back by all the slave nodes respectively;

the second determining module is used for determining an optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes and the initial similarity, wherein the initial similarity is the similarity between a second imaging pattern generated by the preset mask pattern before updating under the target light field and the target pattern;

The third determining module is used for determining whether the optimization result accords with a preset iterative optimization condition;

In a third aspect, the present application provides an exposure apparatus applied to a chip preparation process, comprising: the mask pattern optimizing method comprises a memory and a processor, wherein the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions, so that the mask pattern optimizing method is executed.

In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer instructions for causing a computer to perform the above-described mask pattern optimization method.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic view of a structure of an exposure apparatus applied to a chip manufacturing process according to an embodiment of the present application;

FIG. 2 is a flow chart of a method of optimizing a mask pattern according to an embodiment of the present application;

FIG. 3 is a flow chart of another method of optimizing a mask pattern according to an embodiment of the present application;

FIG. 4 is a schematic diagram of a target pattern and a preset mask pattern in an application scene;

FIG. 5 is a specific workflow diagram of a method of optimizing mask patterns in an application scenario;

FIG. 6 is a block diagram of a mask pattern optimizing apparatus according to an embodiment of the present application;

fig. 7 is a schematic diagram of a hardware configuration of still another exposure apparatus according to an embodiment of the present application.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

It should be understood that the "indication" mentioned in the embodiments of the present application may be a direct indication, an indirect indication, or an indication having an association relationship. For example, a indicates B, which may mean that a indicates B directly, e.g., B may be obtained by a; it may also indicate that a indicates B indirectly, e.g. a indicates C, B may be obtained by C; it may also be indicated that there is an association between a and B.

In the description of the embodiments of the present application, the term "corresponding" may indicate that there is a direct correspondence or an indirect correspondence between the two, or may indicate that there is an association between the two, or may indicate a relationship between the two and the indicated, configured, etc.

In the embodiment of the present application, the "predefining" may be implemented by pre-storing corresponding codes, tables or other manners that may be used to indicate relevant information in devices (including, for example, terminal devices and network devices), and the present application is not limited to the specific implementation manner thereof.

Photolithography is a core technology for integrated circuit fabrication, and due to the continuous development and progress of photolithography, the feature size of integrated circuits that can be processed is continuously reduced, and moore's law is continued. In the photolithography technique, a projection type photolithography machine plays an important role, and a pattern on a mask plate is projected onto a silicon wafer by using a lens or a reflecting mirror, so that transfer of the pattern is realized, and a desired integrated circuit pattern is obtained. In principle, the topological pattern on the mask plate of the projection type photoetching machine is basically consistent with the topological pattern on the silicon wafer, and any defect on the mask can be directly transferred to the corresponding area of the silicon wafer, so that the processing and maintenance cost of the traditional mask plate is higher, and meanwhile, the numerical aperture of a lens is larger and the manufacturing is difficult to improve the imaging resolution.

Furthermore, a calculation holographic photoetching (also called calculation photoetching) is developed, the calculation holographic photoetching is used as a new technology in integrated circuit manufacturing, a mask plate can be synthesized by a computer, then irradiation is carried out through a real light source, and finally, all patterns on the synthesized mask plate can form a desired design pattern after diffraction and interference.

In the calculation holographic lithography, the optimization of the mask pattern (also called mask optimization) is a core problem, and theoretically, if the wavefront obtained after the target pattern is reversely transmitted for a certain distance can be accurately recorded, the target pattern can be well recovered according to the principle of time inversion invariance of light, but because the wavefront is continuous, the more mature mask processing technology is mostly binary amplitude (light transmission or light non-transmission) and binary phase (phase inversion pi), so that a certain optimization method is needed to be adopted to optimize and obtain the mask pattern on the binary amplitude or binary phase mask plate.

In the optimization method of the mask patterns of a plurality of binary masks, the direct search algorithm has the advantages of simplicity in operation, good compatibility and the like, and because each pixel point on the mask patterns is directly evaluated, compared with other optimization methods, the mask optimized by the direct search algorithm is always better in imaging quality or diffraction efficiency. However, the direct search algorithm has large search space, high computational complexity and time dependency in nature, and how to parallelize the direct search algorithm is important for mask optimization of the chip of centimeter magnitude.

In addition, the imaging quality and imaging resolution of the mask are often related to the number of pixels of the mask pattern on the mask, i.e. the mask dimension, and at the same time, for the imaging pattern (also called image plane) of the mask after the light source irradiates, the sampling interval needs to be sufficiently dense to accurately describe the target pattern. Therefore, in order to obtain a large-size high-resolution aerial image, the mask dimension and the image plane dimension are large. In the process of optimizing the mask pattern by adopting the direct search method, the image plane dimension directly influences the time required for evaluating the influence of a single pixel point on the image plane on the mask pattern, and the larger the mask dimension is, the more times required for traversing the whole mask pattern are, and the longer the optimized time is. The direct search method is essentially difficult to be parallel because the mask pattern and the target pattern in the calculation holographic lithography have no local-to-local correspondence, and if the mask pattern is optimized by adopting the original direct search algorithm, the optimization time is difficult to meet the engineering realization, so that improvement is needed for the optimization method for the mask pattern by utilizing the direct search algorithm.

Fig. 1 is a schematic structural diagram of an exposure apparatus for a chip manufacturing process according to an embodiment of the present application, where the exposure apparatus includes a master node 110 and at least one slave node 120, each node is pre-stored with a preset mask pattern and a local pattern in a target pattern, the local patterns stored in the different nodes are different, and the local patterns stored in all the nodes respectively form the target pattern together.

The master node 110 and the slave node 120 are each processors having communication, storage, and data processing functions. Alternatively, the master node 110 and the slave node 120 are respectively graphics processors (Graphics Processing Unit, GPU) independent of each other, or the master node 110 and the slave node 120 are respectively data processors independent of each other, the data processors being composed of at least one central processor (Central Processing Unit, CPU) and at least one GPU. The master node is communicatively connected to each slave node to facilitate subsequent optimization operations of the mask pattern.

Alternatively, the master node 110 and at least one slave node 120 in the exposure apparatus may be disposed in the same computer apparatus of the exposure apparatus, or may be disposed in different computer apparatuses of the exposure apparatus, respectively.

Alternatively, the exposure apparatus of the chip preparation process may be a lithography machine having a computer processing function, or a computer apparatus having a lithography function.

Optionally, after optimization of the mask pattern is completed, a mask fabrication process may also be entered, and a reticle is fabricated based on the optimized mask pattern so that chip fabrication may be performed in the exposure apparatus using the newly fabricated reticle.

Fig. 2 is a flowchart illustrating a method of optimizing a mask pattern for an exposure apparatus for a chip manufacturing process, which is performed by a master node in the exposure apparatus, which may be the master node 110 shown in fig. 1, in an embodiment of the present application. The method may comprise the steps of:

in step 201, in the current iteration period, the pixel value of the target pixel point in the preset mask pattern is updated.

The exposure equipment comprises a master node and at least one slave node, wherein each node is pre-stored with a preset mask pattern and local patterns in a target pattern, the local patterns stored in different nodes are different, and the local patterns respectively stored in all the nodes jointly form the target pattern. The target pattern is a preset pattern to be printed on the wafer in the chip preparation process. The preset mask pattern is a mask pattern to be optimized, and after the optimized preset mask pattern diffracts and interferes light under the irradiation of the target light source when the optimization operation of the preset mask pattern is finished, the formed imaging pattern is very close to the target pattern or is consistent with the target pattern. The target light source is a light source for lithography in an exposure apparatus.

In the current iteration period, the main node selects at least one pixel point in a preset mask pattern based on a direct search algorithm, the selected pixel point is the target pixel point, and the pixel value of the selected pixel point is updated. The embodiment of the application does not limit the direct search algorithm specifically, and can be a linear frequency modulation Scaling (CS) algorithm, a generalized mode search (Generalised Pattern Search, namely GPS) algorithm or a grid self-adaptive direct search (Mesh Adaptive Direct Search, namely MADS) algorithm and the like.

Step 202, determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to a preset similarity calculation rule.

The first local pattern is a local pattern stored in the main node, the sub-imaging pattern is a local imaging pattern in the first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on a pixel value of a target pixel point and a pre-stored preset mask pattern. The main node similarity is used for indicating the proximity degree between the sub-imaging pattern corresponding to the first local pattern and the first local pattern.

After updating the pixel value of the target pixel point, the master node correspondingly updates the preset mask pattern, which means that the imaging pattern of the updated preset mask pattern after being irradiated by the target light source changes, so that before determining the similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern, the master node calculates the imaging pattern of the preset mask pattern after being irradiated by the target light source, namely the first imaging pattern, based on the pixel value of the target pixel point, the diffraction formula of light and the prestored preset mask pattern based on any mask imaging calculation method applied to calculation holographic lithography, and the target light source is a light source used for lithography in exposure equipment.

Although the preset mask pattern and the target pattern have no local-to-local correspondence, there is a one-to-one correspondence between the pixel points between the first imaging pattern and the target pattern. Therefore, after the primary node obtains the first imaging pattern, the sub-imaging pattern corresponding to the first local pattern in the first imaging pattern can be found based on the corresponding relation of the pixel points between the first imaging pattern and the target pattern, and then the primary node can determine the similarity between the first local pattern and the sub-imaging pattern according to a preset similarity calculation rule, and the specific determination step is described in the following embodiments.

And step 203, distributing the pixel values of the target pixel points to different slave nodes, so that the different slave nodes respectively generate new sub-imaging patterns corresponding to the local patterns stored by themselves according to the pixel values of the target pixel points and the preset mask patterns, and determining the slave node similarity between the local patterns stored by themselves and the new sub-imaging patterns.

The master node distributes the pixel value of the target pixel point to different slave nodes, and after receiving the pixel value of the target pixel point, the different slave nodes calculate a first imaging pattern of the preset mask pattern after the target light source irradiates on the basis of the pixel value of the target pixel point, a diffraction formula of light and the preset mask patterns stored in advance by using a mask imaging calculation method which is the same as that of the master node. And finding out a new sub-imaging pattern corresponding to the local pattern stored respectively based on the corresponding relation of the pixel points between the first imaging pattern and the target pattern, and calculating the similarity between the local pattern stored by the rule and the new sub-imaging pattern according to the preset similarity. After the similarity between the local pattern stored by the slave node and the new sub-imaging pattern is determined, the different slave nodes also return the similarity between the local pattern stored by the slave node and the new sub-imaging pattern to the master node respectively.

It should be noted that, the sub-imaging pattern corresponding to the first local pattern stored in the master node and the new sub-imaging pattern corresponding to the local pattern stored in the slave node together form the first imaging pattern. The slave node similarity determined by each slave node is used for indicating the proximity degree between the new sub-imaging pattern corresponding to the local pattern stored by the slave node and the local pattern stored by the slave node.

And 204, obtaining the slave node similarity between the self-stored local patterns and the new sub-imaging patterns fed back by all the slave nodes respectively.

The master node receives the slave node similarity between the self-stored local pattern fed back by the slave node and the new sub-imaging pattern.

Step 205, determining an optimization result of the preset mask pattern in the current iteration period according to the similarity of the master node, the similarity of all the slave nodes respectively determined by the slave nodes and the initial similarity.

The initial similarity is the similarity between the second imaging pattern generated by the preset mask pattern before updating under the target light field and the target pattern. The initial similarity is used to indicate the proximity of the second imaged pattern generated by the pre-set mask pattern before the update to the target pattern under the target light field. The optimized result is the optimized mask pattern determined in the current iteration period.

The master node integrates the similarity of the master node and the similarity of all slave nodes respectively determined by the slave nodes to determine the similarity between the first imaging pattern and the target pattern, and further judges whether the first imaging pattern in the current iteration period is closer to the target pattern according to the similarity between the first imaging pattern and the target pattern and the magnitude relation between the initial similarity, if so, the updated preset mask pattern in the current iteration period is used as an optimization result of the preset mask pattern; if not, taking the pre-set mask pattern before updating in the current iteration period as an optimization result of the pre-set mask pattern.

Step 206, determining whether the optimization result meets a preset iterative optimization condition;

The preset iterative optimization conditions are that the optimization result satisfies at least one of the similarity between the imaging pattern corresponding to the optimization result and the target pattern does not change any more, the similarity between the imaging pattern corresponding to the optimization result and the target pattern is less, and the resolution of the imaging pattern corresponding to the optimization result reaches the preset resolution, so that the optimization operation is stopped. The preset resolution can be set by itself without specific limitation, and it can be understood that the preset iterative optimization conditions can also be set by itself as other conditions to meet various chip preparation scenes.

The specific determination method with small similarity change between the imaging pattern corresponding to the optimization result and the target pattern may be set by itself without specific limitation, for example, when the difference between the similarity between the imaging pattern corresponding to the optimization result and the target pattern and the initial similarity is smaller than a preset value, it may be determined that the similarity change between the first imaging pattern and the target pattern is small, and the preset value may be 0.01 or other self-set values. The method can also be used for monitoring the change of the similarity between the first imaging pattern and the target pattern in each iteration period, and determining that the similarity between the first imaging pattern and the target pattern is smaller when the change amplitude is smaller than a preset threshold value, wherein the preset threshold value can be 0.02 or other self-set values.

In the embodiment of the application, one master node and at least one slave node are arranged in the exposure equipment, each node is pre-stored with a preset mask pattern consistent with other nodes and a local pattern which is different from the other nodes and is part of a target graph, and the common local patterns respectively stored in all the nodes are ensured to form the target pattern, so that the different nodes in the system can be ensured to independently operate according to the data stored by the nodes. In this way, in the current iteration period, the master node updates the pixel value of the target pixel point in the preset mask pattern, and after distributing the pixel value of the target pixel point in the preset mask pattern to different slave nodes, the different slave nodes can independently and parallelly generate a new sub-imaging pattern corresponding to the local pattern stored by themselves according to the pixel value of the target pixel point and the preset mask pattern, and determine the similarity of the local pattern stored by themselves and the new sub-imaging pattern. Meanwhile, the main node also determines the similarity between the sub-imaging pattern corresponding to the first local pattern stored by the main node and the first local pattern according to a preset similarity calculation rule. And obtaining the slave node similarity between the self-stored local patterns fed back by all the slave nodes and the new sub-imaging patterns respectively, so as to determine the optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarity respectively determined by all the slave nodes and the initial similarity. When the optimization result does not accord with the preset iteration optimization condition, entering the next iteration period; or stopping the optimization operation to obtain a final optimization result when the optimization result meets the preset iterative optimization condition. In each iteration, the master node and at least one slave node can simultaneously calculate the similarity between the local pattern stored by the master node and the sub-imaging patterns corresponding to the local pattern in parallel, so that the optimization result of the preset mask pattern is determined, the traditional method for optimizing the mask pattern based on the direct search algorithm is not needed, the calculation time of the similarity between the whole imaging pattern corresponding to the mask pattern and the whole target pattern is shortened to be 1/N, the acceleration ratio is N times of the original acceleration ratio, N is the number of all nodes in the exposure equipment, and the parallel execution of the direct search algorithm is realized, so that the optimization efficiency of the mask pattern is accelerated.

When the optimization of the mask pattern is carried out by utilizing the direct search algorithm, the imaging pattern and the similarity between the imaging pattern and the target pattern are required to be calculated each time when the pixel value on the mask pattern is changed, so that if the imaging pattern is split, each imaging pattern is independently calculated by independent equipment, the parallelization of linear acceleration ratio can be realized, the target pattern can be split, and a plurality of nodes are arranged to independently calculate the similarity between each target pattern and the corresponding sub-imaging pattern. As shown in fig. 3, fig. 3 is a flowchart illustrating another optimization method of mask patterns for an exposure apparatus for a chip manufacturing process according to an embodiment of the present application, which is performed by a master node in the exposure apparatus, which may be the master node 110 shown in fig. 1. The method may comprise the steps of:

in step 301, the target pattern is divided into a preset number of partial patterns.

Wherein the preset number is equal to the number of nodes included in the exposure apparatus. Mode of the master node specifically dividing the target pattern according to the embodiment of the present application is not particularly limited, and the target pattern may be divided into a preset number of partial patterns having different sizes, or may be divided into a preset number of partial patterns having the same size. Preferably, in order to accelerate the optimization efficiency of the mask pattern, the master node may divide the target pattern into a preset number of partial patterns having the same size.

Optionally, before the optimization of the mask pattern starts, the master node further initializes the amplitude and the phase of each pixel in the target pattern and initializes the amplitude and the phase of each pixel in the preset mask pattern to obtain a pixel value P (m, n) of each pixel in the target pattern and a pixel value H (i, j) of each pixel in the preset mask pattern, where (m, n) represents a coordinate value of a pixel in the target pattern and (i, j) represents a coordinate value of a pixel in the preset mask pattern. The slave nodes initialize the amplitude and phase of each pixel point in the preset mask pattern by using the same initialization method as the master node so as to obtain the pixel value H (i, j) of each pixel point in the initialized preset mask pattern, so that the mask pattern can be optimized efficiently.

When the preset mask pattern is the mask pattern in the binary phase mask, the amplitude of each pixel point in the preset mask pattern is a constant, so that the phase of each pixel point can be randomly assigned to be 0 or pi, and the initialization of the preset mask pattern is completed; when the preset mask pattern is a mask pattern in a binary amplitude mask, since the phase of each pixel point in the preset mask pattern is a constant, the amplitude of each pixel point can be randomly assigned to a value indicating light transmission or light non-transmission, so that the initialization of the preset mask pattern is completed.

The embodiment of the application is not particularly limited, and the amplitude and phase values of each pixel point in the target pattern can be designed in advance.

In step 302, other local patterns except the first local pattern are distributed to different slave nodes, respectively.

The first partial pattern is any one of a preset number of partial patterns. The master node distributes the other partial patterns than the first partial pattern to different slave nodes, respectively. For example, as shown in fig. 4, taking an example in which one master node and 3 slave nodes are included in the exposure apparatus, the master node divides the target pattern into four partial patterns of P0, P1, P2, and P3, the master node stores P0, distributes P1 to the slave node 0, distributes P2 to the slave node 1, and distributes P3 to the slave node 2. It can be understood that when the master node distributes the coordinate values and the pixel values of the pixel points in other local patterns to different slave nodes, the slave nodes can acquire the coordinate values and the pixel values of each pixel point in the local pattern distributed to the slave nodes, so that the similarity between the local pattern stored in the slave nodes and the new sub-imaging pattern can be conveniently determined.

Before updating a preset mask pattern, the master node divides the target pattern into local patterns with the same number as the number of nodes included in the exposure device, and distributes other local patterns except the first local pattern to different slave nodes respectively, so that each slave node can acquire one local pattern, and further, when the mask pattern is optimized subsequently, all nodes of the exposure device can continuously and parallelly calculate the similarity between the local pattern stored by the master node and the corresponding sub-imaging pattern, thereby saving the time when the mask pattern is optimized by using a direct search algorithm and submitting the optimization efficiency of the mask source.

In step 303, in the current iteration period, the pixel value of the target pixel point in the preset mask pattern is updated.

Please refer to step 201 in the embodiment shown in fig. 2 in detail, which is not described herein.

Optionally, in consideration of the fact that the pixel values of the pixel points of the mask patterns on the different types of masks need to be updated in the practical application process, in step 303, updating the pixel values of the target pixel points in the preset mask pattern in the current iteration period may include:

The direct search algorithm may be a CS algorithm, a GPS algorithm, a MADS algorithm, or the like. When the preset mask pattern is the mask pattern in the binary phase mask, the phase change operation performed on the target pixel point is to turn over the current phase of the target pixel point, taking the phase of the pixel point in the binary phase mask as 0 or pi as an example, and if the current phase of the current target pixel point is 0, turning over the phase to pi; if the current phase of the current target pixel point is pi, the phase is flipped to 0. For example, when the preset mask pattern is a mask pattern in a binary phase mask, the master node selects a certain pixel point (i _p ，j _p ) The phase is changed to pi if the original phase is 0, and to 0 if the original phase is pi.

When the preset mask pattern is the mask pattern in the binary amplitude mask, performing amplitude change operation on the target pixel as turning over the current amplitude of the target pixel, taking the amplitude of the pixel in the binary amplitude mask as the pixel to transmit light or the pixel to not transmit light as an example, and turning over the amplitude as the pixel to transmit light if the current amplitude of the current target pixel is the pixel to not transmit light; if the current amplitude of the current target pixel point is light transmission of the pixel point, the amplitude is turned over to be light-proof of the pixel point.

It should be noted that, the specific value of the phase of the pixel point in the binary phase mask is only two possible, in the embodiment of the present application, the value of the phase may be 0 or pi, but is not limited to 0 or pi, but may be 0 or pi/2, or other values that are set by themselves, such as pi/2 or pi. The specific value of the amplitude of the pixel in the binary amplitude mask may be similar, and is not limited to the transparent or opaque pixel, but may be any other value that may be set by itself, such as transparent or semi-transparent pixel.

Step 304, determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to a preset similarity calculation rule.

Please refer to step 202 in the embodiment shown in fig. 2 in detail, which is not described herein.

Optionally, in order to further increase the optimization efficiency of the mask pattern, step 304 of determining, according to a preset similarity calculation rule, a main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern may include:

acquiring a pixel value of each pixel point in the first local pattern;

And acquiring a pixel value of each pixel point of the sub-imaging pattern corresponding to the first local pattern.

The master node obtains the pixel value P of each pixel point in the first local pattern after initializing the target pattern _q (m _q ，n _q ). The main node obtains the pixel value of each pixel point in the first imaging pattern after calculating the first imaging pattern, and then based on the corresponding relation of the pixel points between the first imaging pattern and the target pattern, the pixel value E of each pixel point of the sub imaging pattern corresponding to the first local pattern can be obtained when the sub imaging pattern corresponding to the first local pattern in the first imaging pattern is found out _q (x _q ，y _q ) Wherein q represents the qth partial pattern in the target pattern, which may be referred to herein as the first partial pattern, (m _q ，n _q ) A coordinate value (x) representing a pixel point in the first partial pattern _q ，y _q ) And the coordinate value of a pixel point in the sub-imaging pattern corresponding to the first local pattern is represented.

The master node substitutes the pixel value of each pixel point in the first local pattern and the sub-imaging pattern corresponding to the first local pattern into the formulaAnd calculating the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern. In the formula loss _i Representing the similarity between the ith local pattern and the corresponding sub-imaging pattern of the ith local pattern in the target pattern,/for the target pattern>Representing the pixel value of the j-th pixel point in the sub-imaging pattern corresponding to the first local pattern,/and>and representing the pixel value of the j-th pixel point in the first local pattern, and z represents the total number of all the pixel points in the first local pattern.

And step 305, distributing the pixel values of the target pixel points to different slave nodes, so that the different slave nodes respectively generate new sub-imaging patterns corresponding to the local patterns stored by themselves according to the pixel values of the target pixel points and the preset mask patterns, and determining the slave node similarity between the local patterns stored by themselves and the new sub-imaging patterns.

Please refer to step 203 in the embodiment shown in fig. 2 in detail, which is not described herein.

It should be noted that, the specific method of the slave node similarity between the local pattern stored by the slave node and the new sub-imaging pattern is similar to that of the master node, and the pixel value of each pixel point in the local pattern stored by the slave node and the new sub-imaging pattern is substituted into the formula Self-memory of the middle calculationSlave node similarity between the stored local pattern and the new sub-imaged pattern. In addition, for the slave nodes, before the new sub-imaging pattern corresponding to the local pattern stored by the slave nodes, the pixel value of the selected pixel point on the preset mask pattern sent by the master node is received, and after the similarity between the local pattern stored by the slave nodes and the corresponding new sub-imaging pattern is calculated, each slave node only needs to transmit the corresponding determined similarity to the master node, so that the optimized communication cost of the mask pattern is low.

And step 306, obtaining the slave node similarity between the self-stored local patterns and the new sub-imaging patterns fed back by all the slave nodes respectively.

Please refer to step 204 in the embodiment shown in fig. 2 in detail, which is not described herein.

Step 307, determining an optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes and the initial similarity.

Please refer to step 205 in the embodiment shown in fig. 2 in detail, which is not described herein.

Optionally, in step 307, the determining an optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes, and the initial similarity may include the following steps 3071 to 3072:

In step 3071, a total similarity between the first imaging pattern and the target pattern is determined according to the master node similarity and the slave node similarities respectively determined by all the slave nodes.

After receiving the slave node similarities fed back by different slave nodes, the master node accumulates the master node similarities determined by the master node and all the slave node similarities, and determines the accumulated sum value as the total similarity between the first imaging pattern and the target pattern in the current iteration period, wherein the specific accumulated formula is as followsWherein loss represents the total phase between the first imaging pattern and the target pattern in the current iteration periodSimilarly, N is the number of all nodes in the exposure apparatus.

It should be noted that, the initial similarity is used as the similarity between the second imaging pattern and the target pattern generated by the preset mask pattern before updating in the target light field, and when the current iteration period is the first iteration period, the initial similarity is the total similarity between the initialized preset mask pattern and the initialized target pattern, and the specific calculation method of the initial similarity is similar to the calculation method of the total similarity between the first imaging pattern and the target pattern, and only the step of updating the preset mask pattern by the main node is omitted, which is not repeated here.

Optionally, in the first iteration period, the master node may also independently calculate the initial similarity, specifically, after initializing the target pattern and the preset mask pattern, the master node calculates, according to a diffraction formula of light, an imaging pattern projected after the irradiation of the target light source in the initialized preset mask pattern, thereby obtaining a pixel value E of each pixel point in the imaging pattern ₀ (x, y), wherein (x, y) represents the coordinate values of the pixel points in the imaging pattern. Then substituting the pixel value of each pixel point in the imaging pattern and the pixel value of each pixel point in the target pattern into the formulaWherein the initial similarity between the imaging pattern and the target pattern is calculated, wherein +.>Representing initial similarity,/->Representing the pixel value of the g-th pixel point in the imaging pattern projected after the irradiation of the target light source in the initialized preset mask pattern,/for the imaging pattern>The pixel value of the g-th pixel point in the target pattern is represented, and f represents the total number of all the pixel points in the target pattern.

Step 3072, determining an optimization result according to the initial similarity and the total similarity.

The main node compares the total similarity with the initial similarity to determine whether the first imaging pattern is closer to the target pattern, and if so, the updated preset mask pattern in the current iteration period is used as an optimization result of the preset mask pattern; if not, taking the pre-set mask pattern before updating in the current iteration period as an optimization result of the pre-set mask pattern.

The master node determines the total similarity between the first imaging pattern and the target pattern according to the similarity of the master node and the similarity of the slave nodes respectively determined by all the slave nodes, and the master node is not required to directly calculate the total similarity between the whole imaging pattern corresponding to the mask pattern and the whole target pattern. Each node can determine a similarity, and then when the total similarity is determined, the calculation efficiency of the total similarity is enabled to be hooked with the number of the master nodes and the slave nodes, parallel calculation of the total similarity is achieved, and when the direct search algorithm is used for optimizing the mask pattern, parallelization design of the direct search algorithm is achieved, and therefore optimization efficiency of the mask pattern is accelerated.

Optionally, in step 3072, determining the optimization result according to the initial similarity and the total similarity may include:

When the initial similarity is greater than or equal to the total similarity, meaning that the similarity difference between the second imaging pattern and the target pattern is greater than the similarity between the first imaging pattern and the target pattern, the first imaging pattern is closer to the target pattern, and the updated preset mask pattern corresponding to the first imaging pattern is determined as an optimization result. The initial similarity loss determined in the previous iteration period is also determined _e Is updated to the value of the total similarity loss determined in the current iteration period, namely loss _e =loss, so that the next iteration cycle determines the optimization result. And determining the updated preset mask pattern as an optimized result，

Optionally, when the initial similarity is smaller than the total similarity, the updated preset mask pattern in the current iteration period is restored to the mask pattern acquired in the previous iteration period, the mask pattern acquired in the previous iteration period is used as the original mask pattern of the next iteration period, and the optimization operation of the next iteration period is performed to acquire a new iteration optimization result again;

The pre-configured initial mask pattern is the initialized pre-configured mask pattern. The mask pattern obtained in the previous iteration cycle is the optimization result determined in the previous iteration cycle, namely the preset mask pattern before updating in the current iteration cycle. When the initial similarity is smaller than the total similarity, the main node restores the pixel value of the updated target pixel point to the pixel value of the target pixel point before the updating of the preset mask pattern in the current iteration period. For example, when the preset mask pattern is a mask pattern in a binary phase mask, in the current iteration period, a pixel point (i) _p ，j _p ) Pixel value H (i) _p ，j _p ) =0, and the pixel value H (i _p ，j _p ) =pi, initial similarity loss determined in the previous iteration cycle _e Less than the total similarity loss determined in the current iteration period, the updated pixel point (i _p ，j _p ) Is reduced from pi to 0.

Step 308, determining whether the optimization result meets a preset iterative optimization condition;

Please refer to step 206 in the embodiment shown in fig. 2 in detail, which is not described herein.

Alternatively, the master node may divide the target pattern into a plurality of local patterns by taking the number of the slave nodes as a reference instead of storing the first local pattern, and distribute the plurality of local patterns to different slave nodes respectively, so that the different slave nodes determine the slave node similarity between the local pattern stored by the master node and the new sub-imaging pattern, and then the master node obtains the slave node similarity fed back by all the slave nodes respectively, so as to determine the total similarity between the first imaging pattern and the target pattern according to the slave node similarity determined by all the slave nodes, and further determine the optimization result. It should be noted that, when the master node does not store the first local pattern, the master node does not need to calculate the similarity between the first local pattern and the sub-imaging pattern corresponding to the first local pattern, where the number of slave nodes in the exposure apparatus should be greater than or equal to 2.

In an application scenario, as shown in fig. 5, taking a preset mask pattern as a mask pattern on a binary phase mask, an exposure device includes a master node and four slave nodes, where the master node does not store a first local pattern, and a specific workflow of an optimization method of the mask pattern may include the following steps:

step 1, initializing the amplitude and phase of pixel points in a mask pattern on a binary phase mask, randomly assigning the phase of each pixel point on the binary phase mask to 0 or pi to obtain H (i, j), as shown in fig. 4, and setting the pixel value P (m, n) of each pixel point in a target pattern, as shown in fig. 4, as shown in initialization H and P in fig. 5;

step 2, calculating an imaging pattern E of the initialized mask pattern according to a diffraction formula ₀ (m, n), comparison E ₀ (m, n) and P (m, n), calculating the loss(this loss is the total similarity in the embodiment of the present application, here the initial similarity calculated in the first iteration cycle), i.e. in FIG. 5, E is calculated from H ₀ From E ₀ And P calculate loss _e ；

Step 3, splitting the target pattern into N partial patterns spatiallyObtaining a local pattern P _i Where i takes 0,1,2,..n-1, represents the i-th partial pattern, N being the number of all slave nodes in the exposure apparatus. In the application scene, the partial patterns are split into 4 partial patterns as shown in fig. 4, so as to obtain partial patterns P0, P1, P2 and P3, and the partial patterns P0 are distributed to slave nodes 0, P1 to slave nodes 1, P2 to slave nodes 2 and P3 to slave node 3;

Step 4, selecting a pixel point (i) _p ，j _p )，H(i _p ，j _p ) =0, and changing its phase (pi if the original phase is 0 ), H (i) _p ，j _p ) =0 to H (i _p ，j _p ) =pi; the pixel value H (i _p ，j _p ) Distributed to different slave nodes; different slave nodes respectively calculate own stored local patterns P on respective graphics processors in parallel _i Corresponding sub-imaging pattern E _i And is composed of E _i P _i Respectively calculating self-stored local patterns P _i And corresponding sub-imaging pattern E _i Loss between(the loss is the similarity of slave nodes in the embodiment of the application), and loss is obtained _i The slave node will lose _i Transmitting to the master node; after the master node receives, calculate the total loss +.>；

Step 5, comparing the loss, if loss<loss _e Receiving the change of the pixel point and updating loss _e =loss, go back to step 4 and continue, otherwise reject the change, add the pixel value H (i _p ，j _p ) Return pi to original value H (i _p ，j _p ) =0, go back to step 4 and continue until the requirement is met (loss no longer changes or changes are smaller), output the holographic mask data file (i.e. the mask pattern corresponding to the optimization result), and stop the optimization operation.

Through the steps, the parallelization design of the binary direct search algorithm can be realized, the acceleration ratio is N times, namely the number of blocks divided by the image plane is N, and the calculation time is changed into 1/N.

FIG. 6 is a block diagram of an apparatus for optimizing a mask pattern applied to an exposure device of a chip manufacturing process, the exposure device including a master node and at least one slave node, each node having a preset mask pattern and a local pattern in a target pattern pre-stored therein, the local patterns stored in the different nodes being different, the local patterns stored in all nodes respectively forming the target pattern together; the apparatus is operated on a master node, and the mask pattern optimizing apparatus includes:

an updating module 610, configured to update a pixel value of a target pixel point in a preset mask pattern in a current iteration period;

a first determining module 620, configured to determine, according to a preset similarity calculation rule, a main node similarity between a sub-imaging pattern corresponding to a first local pattern and the first local pattern, where the first local pattern is a local pattern stored in the main node, the sub-imaging pattern is a local imaging pattern in the first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on a pixel value of a target pixel and a pre-stored preset mask pattern;

the distributing module 630 is configured to distribute the pixel values of the target pixel points to different slave nodes, so that the different slave nodes generate new sub-imaging patterns corresponding to the local patterns stored in the slave nodes according to the pixel values of the target pixel points and the preset mask patterns, and determine the slave node similarity between the local patterns stored in the slave nodes and the new sub-imaging patterns;

An obtaining module 640, configured to obtain slave node similarities between the self-stored local patterns and the new sub-imaging patterns fed back by all the slave nodes respectively;

a second determining module 650, configured to determine an optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes, and the initial similarity, where the initial similarity is a similarity between a target pattern and a second imaging pattern generated by the preset mask pattern before updating under the target light field;

a third determining module 660, configured to determine whether the optimization result meets a preset iterative optimization condition;

In some optional embodiments, before updating the pixel value of the target pixel point in the preset mask pattern in the current iteration period, the optimization device of the mask pattern further includes:

a dividing module for dividing the target pattern into a preset number of partial patterns, the preset number being equal to the number of nodes included in the exposure apparatus;

And the distribution module is also used for distributing other local patterns except the first local pattern to different slave nodes respectively.

In some alternative embodiments, the update module includes:

the changing unit is used for selecting a target pixel point in the preset mask pattern to carry out phase changing operation based on a direct searching algorithm when the preset mask pattern is the mask pattern in the binary phase mask plate so as to change the pixel value of the target pixel point;

In some alternative embodiments, the first determining module includes:

an obtaining unit, configured to obtain a pixel value of each pixel point in the first local pattern;

the first determining unit is used for determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to the pixel values of all the pixel points in the first local pattern and the pixel values of all the pixel points in the sub-imaging pattern corresponding to the first local pattern.

In some alternative embodiments, the second determining module includes:

a second determining unit, configured to determine a total similarity between the first imaging pattern and the target pattern according to the master node similarity and the slave node similarities respectively determined by all the slave nodes;

and the third determining unit is used for determining an optimization result according to the initial similarity and the total similarity.

In some alternative embodiments, the third determining unit comprises:

and the updating subunit is used for updating the value of the initial similarity into the value of the total similarity when the initial similarity is larger than or equal to the total similarity, and determining the updated preset mask pattern as an optimization result.

In some alternative embodiments, when the initial similarity is less than the total similarity, the optimizing means of the mask pattern further comprises:

the restoring module is used for restoring the updated preset mask pattern in the current iteration period into the mask pattern acquired in the previous iteration period, taking the mask pattern acquired in the previous iteration period as the original mask pattern of the next iteration period, and entering the optimization operation of the next iteration period so as to acquire a new iteration optimization result again;

Further functional descriptions of the above respective modules and units are the same as those of the above corresponding embodiments, and are not repeated here.

The mask pattern optimizing means in this embodiment is presented in the form of functional units, here referred to as ASIC (Application Specific Integrated Circuit ) circuits, processors and memories executing one or more software or fixed programs, and/or other devices that can provide the above described functions.

The embodiment of the application also provides exposure equipment which is applied to the chip preparation process and is provided with the optimization device of the mask pattern shown in the figure 6.

Referring to fig. 7, fig. 7 is a schematic structural diagram of another exposure apparatus according to an alternative embodiment of the present application, as shown in fig. 7, the exposure apparatus includes: one or more processors 10, memory 20, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are communicatively coupled to each other using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions executed within the exposure apparatus, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display apparatus coupled to the interface. In some alternative embodiments, multiple processors and/or multiple buses may be used, if desired, along with multiple memories and multiple memories. Also, multiple exposure apparatuses may be connected, each providing a portion of the necessary operations (e.g., as a server array, a set of blade servers, or a multiprocessor system). One processor 10 is illustrated in fig. 7.

The processor 10 may be a central processor, a network processor, or a combination thereof. The processor 10 may further include a hardware chip, among others. The hardware chip may be an application specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general-purpose array logic, or any combination thereof.

Wherein the memory 20 stores instructions executable by the at least one processor 10 to cause the at least one processor 10 to perform the methods shown in implementing the above embodiments.

The memory 20 may include a storage program area that may store an operating system, at least one application program required for functions, and a storage data area; the storage data area may store data created according to the use of the exposure apparatus, and the like. In addition, the memory 20 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, memory 20 may optionally include memory located remotely from processor 10, which may be connected to the exposure apparatus via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

Memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk, or solid state disk; the memory 20 may also comprise a combination of the above types of memories.

The exposure apparatus further comprises a communication interface 30 for the exposure apparatus to communicate with other apparatuses or communication networks.

The embodiments of the present application also provide a computer readable storage medium, and the method according to the embodiments of the present application described above may be implemented in hardware, firmware, or as a computer code which may be recorded on a storage medium, or as original stored in a remote storage medium or a non-transitory machine readable storage medium downloaded through a network and to be stored in a local storage medium, so that the method described herein may be stored on such software process on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, a solid state disk or the like; further, the storage medium may also comprise a combination of memories of the kind described above. It will be appreciated that a computer, processor, microprocessor controller or programmable hardware includes a storage element that can store or receive software or computer code that, when accessed and executed by the computer, processor or hardware, implements the methods illustrated by the above embodiments.

Although embodiments of the present application have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the application, and such modifications and variations fall within the scope of the application as defined by the appended claims.

Claims

1. The optimization method of the mask pattern is characterized by being applied to exposure equipment of a chip preparation process, wherein the exposure equipment comprises a master node and at least one slave node, each node is prestored with a preset mask pattern and local patterns in a target pattern, the local patterns stored in different nodes are different, and the local patterns respectively stored in all the nodes jointly form the target pattern; the method is performed by the master node, the method comprising:

updating the pixel value of the target pixel point in the preset mask pattern in the current iteration period;

determining the similarity of a sub-imaging pattern corresponding to a first local pattern and a main node between the first local pattern according to a preset similarity calculation rule, wherein the first local pattern is a local pattern stored by the main node, the sub-imaging pattern is a local imaging pattern in a first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on the pixel value of the target pixel point and a pre-stored preset mask pattern;

Distributing the pixel value of the target pixel point to different slave nodes so that the different slave nodes respectively generate a new sub-imaging pattern corresponding to the local pattern stored by themselves according to the pixel value of the target pixel point and the preset mask pattern, and determining the slave node similarity between the local pattern stored by themselves and the new sub-imaging pattern;

obtaining the similarity of slave nodes between the self-stored local patterns fed back by all the slave nodes and the new sub-imaging patterns respectively;

determining an optimization result of a preset mask pattern in a current iteration period according to the master node similarity, the slave node similarities respectively determined by all slave nodes and the initial similarity, wherein the initial similarity is the similarity between a second imaging pattern generated by the preset mask pattern before updating under a target light field and the target pattern;

when the optimization result is determined to be not in accordance with the preset iterative optimization condition, entering a next iterative period;

or stopping the optimization operation when the optimization result is determined to be in accordance with the preset iterative optimization condition.

2. The method of claim 1, wherein before updating the pixel value of the target pixel point in the preset mask pattern in the current iteration period, the method further comprises:

dividing a target pattern into a preset number of partial patterns, the preset number being equal to the number of nodes included in the exposure apparatus;

and distributing other local patterns except the first local pattern to different slave nodes respectively.

3. The method according to claim 1 or 2, wherein updating the pixel value of the target pixel point in the pre-stored preset mask pattern comprises:

when the preset mask pattern is a mask pattern in a binary phase mask, selecting a target pixel point in the preset mask pattern to perform phase change operation based on a direct search algorithm so as to modify the pixel value of the target pixel point;

or when the preset mask pattern is the mask pattern in the binary amplitude mask, selecting a target pixel point in the preset mask pattern to carry out amplitude change operation based on a direct search algorithm so as to modify the pixel value of the target pixel point.

4. The method according to claim 1 or 2, wherein determining the main node similarity between the sub-imaging pattern corresponding to the first local pattern and the first local pattern according to a preset similarity calculation rule includes:

Acquiring a pixel value of each pixel point in the first local pattern;

5. The method according to claim 1, wherein determining the optimization result of the preset mask pattern in the current iteration period according to the master node similarity, the slave node similarities respectively determined by all the slave nodes, and the initial similarity comprises:

determining the total similarity between the first imaging pattern and the target pattern according to the master node similarity and the slave node similarity respectively determined by all the slave nodes;

and determining the optimization result according to the initial similarity and the total similarity.

6. The method of claim 5, wherein said determining said optimization result from said initial similarity and said total similarity comprises:

And when the initial similarity is greater than or equal to the total similarity, updating the value of the initial similarity to the value of the total similarity, and determining the updated preset mask pattern as the optimization result.

7. The method of claim 6, wherein when the initial similarity is less than the total similarity, the method further comprises:

or when the current iteration period is the first iteration period, restoring the updated preset mask pattern in the current iteration period into a preset initial mask pattern.

8. The device is characterized by being applied to exposure equipment of a chip preparation process, wherein the exposure equipment comprises a master node and at least one slave node, each node is prestored with a preset mask pattern and local patterns in a target pattern, the local patterns stored in different nodes are different, and the local patterns respectively stored in all the nodes jointly form the target pattern; the apparatus operating on the master node, the apparatus comprising:

the first determining module is used for determining the similarity of a main node between a sub-imaging pattern corresponding to a first local pattern and the first local pattern according to a preset similarity calculation rule, wherein the first local pattern is a local pattern stored by the main node, the sub-imaging pattern is a local imaging pattern in the first imaging pattern, and the first imaging pattern is an imaging pattern pre-generated based on the pixel value of the target pixel point and a pre-stored preset mask pattern;

the distribution module is used for distributing the pixel value of the target pixel point to different slave nodes so that the different slave nodes respectively generate a new sub-imaging pattern corresponding to the local pattern stored by the slave nodes according to the pixel value of the target pixel point and the preset mask pattern, and determine the slave node similarity between the local pattern stored by the slave nodes and the new sub-imaging pattern;

the acquisition module is used for acquiring the slave node similarity between the self-stored local patterns fed back by the slave nodes and the new sub-imaging patterns respectively;

The second determining module is used for determining an optimization result of a preset mask pattern in a current iteration period according to the similarity of the master node, the similarity of all the slave nodes respectively determined by the slave nodes and the initial similarity, wherein the initial similarity is the similarity between a second imaging pattern generated by the preset mask pattern before updating under a target light field and the target pattern;

a third determining module, configured to determine whether the optimization result meets a preset iterative optimization condition;

9. An exposure apparatus applied to a chip preparation process, comprising:

a memory and a processor in communication with each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform the method of optimizing a mask pattern according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon computer instructions for causing a computer to execute the optimization method of a mask pattern according to any one of claims 1 to 7.