JPH1185826A - Plotting data processing system and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the efficiency of processing by reducing the transfer load of a network between a client and a server. SOLUTION: At a data processing system for a plotting device, provided with a parallel processing function by plural work stations connected through a network 13 or the plural servers 12 of a client/serve model by a personal computer, data of the entire chips is transferred in advance from a client to the servers 12, a client 11 transfers designated information of a part desired to process to the servers 12, the servers 12 extract graphic data of a pertinent part from the entire data already transferred to execute graphic arithmetical processing, etc., and transfers the result to the client 11, and the client 11 merges the result to prepare a final data file for plotting 15.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a drawing data processing system and a drawing data processing method, and more particularly to an LSI.
The present invention relates to a drawing data processing system for generating LSI pattern electron beam drawing apparatus data from design pattern data and a method for processing drawing data.

[0002]

2. Description of the Related Art In recent years, LSI patterns have become finer and more complex, and the conversion processing time for creating data for an electron beam lithography apparatus from LSI design pattern data has become enormous. As one method of reducing the processing time, a parallel processing system of a client-server model using a plurality of workstation machines or personal computers connected via a network is used.

This parallel processing system has a configuration as shown in FIG. 15, and a processing flowchart is as shown in FIG. In FIG. 15, 11 is a client machine,
12-1, 12-2... Are server machines, 13 is a network, 14 is design data, 15 is drawing data, 1
Numeral 6 denotes data of the entire chip, and 16-1 and 16-2 denote respective portions divided from the data 16 of the entire chip.

[0004] The operation of this system will be described with reference to the processing flowchart of FIG. First, from the design data 14, the LSI
The design pattern data is input, the hierarchy is expanded (step 301), and the entire data is divided into small areas (step 302). Each of the divided data is transferred to another machine 12 (step 303), a graphic operation is performed in the machine 12, and the result is transferred to the original machine (step 304). The results transferred from each machine are merged (step 305), and when the processing of all the divided areas is completed (step 306), the final drawing apparatus data 15 is output. Below, receive the divided data,
The machine that performs the arithmetic processing is the server 12, and the machine that divides the entire data and transfers it to the server 12 is the client 1.
It is referred to as 1.

Usually, this division unit is often used as a drawing unit in a drawing apparatus. However, since the division unit is small compared to the whole chip size, the number of divided areas becomes enormous. According to the huge number of areas, the number of times of communication exchange between the client and the server becomes enormous, so that there is a problem that a network transfer load becomes extremely large and parallel processing efficiency is reduced.

Further, the client performs a process of transferring data to the server and a process of receiving data from the server and adding it to output data. However, since this process takes time, the client constantly performs some process. In such a case, even if the server transfers the processing result to the client, the server cannot immediately receive the next processing data, which causes a waiting time. Therefore, there is a problem that the parallel processing efficiency is limited by the load efficiency of the client. It is desired to solve such a problem.

Conventionally, when it is desired to process LSI pattern data, it is necessary to prepare software as needed and execute it, or use a commercially available CA.
This was addressed by using a D tool or a verification tool. However, in the case of the former, data processing different from the previous time is necessary, and if the previously created software can not handle it, it must be handled by adding functions to the software or creating new software In the meantime, while repeating such a thing, maintenance management became difficult and the frequency of increase increased, causing a problem that the burden became large. On the other hand, in the latter case, since it is not a tool dedicated to data processing, there have been problems such as a slow execution time and some data processing that cannot be handled.

In particular, with the recent evolution of the generation of LSIs, higher densities of semiconductor circuits are required.
The pattern data size is also increasing. In order to perform such data processing, it is troublesome and time-consuming to execute software for the type of processing and to load the data into a CAD tool or a verification tool each time.
und Time: the time required for development). The solution of this problem is also an important problem for creating data for an electron beam lithography apparatus.

Further, there is a problem related to clip (extraction) processing. In the process of creating mask drawing data from LSI design data for drawing a mask used in semiconductor manufacturing, as described above, the entire data is reduced to a small area for the purpose of reducing the drawing data. To divide. That is, a method of clipping LSI design data in a specific area and dividing the data into a plurality of drawing data files is performed.

Here, the necessity of clip processing will be considered. For example, when a plurality of patterns of the same LSI chip are mounted on a mask, it is wasteful to process the whole as it is from the viewpoint of data processing time and the size of drawing data to be created. In such a case, the pattern data other than the LSI chip portion and the pattern data for one LSI chip are separated, a pattern file for each drawing is created, and the pattern of the LSI chip is repeatedly drawn at the time of mask drawing. Thus, the pattern of the entire mask is drawn. By doing so, in the case of a mask on which the same LSI of four chips is mounted, the processing time for drawing data creation and the drawing data can be substantially reduced to about ４ compared to the method of batch processing the whole. I can do it.

As a similar example, consider a memory. In the case of a memory, many areas in the chip are filled with the same pattern. Therefore, as long as mask drawing overhead due to repeated drawing is not a problem,
The repeat unit of the pattern is set, and the same LSI
A method is used in which drawing data is created in the same manner as a chip, and the same pattern portion in the memory is repeatedly drawn by a set repeating unit. As described above, clip processing is used to save the drawing data creation time and the drawing data size.

For other purposes, for example, a large-scale L
If you need data focusing on a specific part of SI,
Since it is inefficient to process the whole, data is created by extracting (clipping) only necessary parts.

In this clip processing procedure, it is determined whether each input data is inside or outside the clip area. In the case of clip processing that leaves the outside of the designated area, data inside the area is discarded, and outside data is left.
With respect to a graphic that intersects with the area rectangle, clip processing such as cutting out the graphic is performed.

When the input data is a repeated expression,
In the whole repetition, it is determined whether it is inside or outside the above-mentioned repetition area, and the same processing is performed. Is performed. FIG. 17 shows an example of a clip of a figure having repetition information.

In this example, all the repetitive graphics are developed, and the inside / outside / intersection determination for the clip region is performed for all 48 graphics. Here, considering the clipping process of the repetitive pattern portion in the memory shown in the above example, the outermost peripheral portion of the memory generally has a different shape from the internal pattern due to communication with an external pattern. In many cases, it is necessary to remove the innermost portion while leaving the outermost portion. Further, when the inner pattern excluding the outermost periphery does not have a size of an integral multiple of the repeating unit, the surplus portion must be left without being clipped similarly to the outermost peripheral portion.

If the memory has a large scale, the repetition of the pattern portion to be left without clipping is developed, so that the processing time required for clipping becomes long, and the output data size after clipping becomes huge, This affects the processing time until the drawing data is created in the subsequent processing. Therefore, the effect of shortening the original processing time cannot be expected.

The number of figures is checked by simple calculation. 16
It is assumed that the Mb memory is a simple repetition in units of 1 Mb, and clip processing is performed at 16 locations in units of 1 Mb. Since the purpose is to leave the outermost periphery, the repetition of the 1 Mb pattern intersects with the clip region rectangle, and the process of determining whether the pattern of all the bits developed by repetition is inside or outside the clip region Becomes Also, assuming that the data amount of the clipping result is such that each bit of 1 Mb is arranged in a simple rectangular array, since 1 Mb is 1024 ² , the number of bits in the outermost peripheral portion is 4092. That is,
One bit of data having a repetition instruction of 1 Mb is
After clipping the outermost peripheral portion, the data becomes 4092 bits of data without a repeat instruction. Solving such a clip-related problem and performing an efficient process is also an important problem in creating data for an electron beam drawing apparatus.

[0018]

As described above, there are several problems in the data generating apparatus and method for the electron beam lithography apparatus. The first problem is to reduce the network transfer load between the client and the server. In the conventional client-server model parallel processing system,
The entire pattern data was divided, and the divided data was sent to the server each time to perform the arithmetic processing.However, the number of divided data became enormous, the communication volume between the client and the server increased, and the network transfer load was reduced. There is a problem that the processing efficiency increases and the processing efficiency decreases.

The second problem is to deal with the addition or improvement of software functions. Conventionally, if you want to process the LSI pattern data, you need to create software as needed and execute it, or use a commercially available CAD tool or verification tool. In the former case, However, if the previously created software cannot be used, it is necessary to add functions to the software or create new software, which will increase the burden on work and maintenance. In the latter case, since the tool is not a tool dedicated to data processing, there have been problems such as a slow execution time and some data processing that cannot be handled.

The third problem is related to the division of a graphic having repetition information, and the processing of a graphic having repetition information, which has been conventionally performed mechanically, is made as efficient as possible by storing the repetition information as much as possible. There was a problem to try to.

The present invention solves these problems, reduces the network transfer load between a client and a server, facilitates processing for adding or improving software functions, and efficiently divides a graphic having repeated information. It is an object to realize a drawing data creation system and a drawing data creation method.

[0022]

To achieve the above object, the present invention provides a parallel processing function of a plurality of servers of a client / server model including a plurality of workstations or personal computers connected by a network. In a drawing data processing system that processes pattern data to create drawing data and a method of processing such drawing data, the entire pattern data is transferred to the plurality of servers in advance, and the client sends the pattern data to the server. The server transfers the specification information of the part to be processed, the server extracts the graphic data of the part specified by the specification information from the transferred entire pattern data, performs a graphic operation process, and outputs the result to the client. And the client merges the result to the final drawing data file. Characterized by creating a Le.

[0023] Further, a drawing function is provided that has a parallel processing function of a plurality of servers of a client / server model including a plurality of workstations or personal computers connected via a network, and that processes drawing data to generate drawing data. In a data processing system and such a method of processing drawing data, in a process of reading data by software and transferring data output individually, a selection condition and a change condition for a basic information unit are designated, and thus the selection condition is matched. The basic information unit to be changed is changed according to the contents of the change condition.

Further, in a drawing data processing system for processing pattern data to generate drawing data and a method for processing such drawing data, a pattern data group including repetition information, which is expressed as a file on a computer, is used to generate a pattern data group. When cutting and extracting a region of interest, processing is performed so that the original repetition information is stored as much as possible in the extraction region, and the repetition information is processed in repetition units.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a drawing data creation system and a drawing data creation method according to the present invention will be described in detail with reference to the accompanying drawings.

First, the first invention of the present invention addresses the problem of reducing the network transfer load between a client and a server as follows. That is, first, the entire data is transferred from the client to the server, and thereafter, information on the position and size of the storage area relating to the portion to be processed by each server is specified. Each server is
The graphic data of the designated portion is extracted from the previously transferred whole data, arithmetic processing is performed, and the result is transferred to the client. The client performs a process of merging the data sent from each server with the output data, and the final output data is completed when all the processes are completed.

In other words, the client does not transfer the actual graphic data to each server in the individual processing, but only transfers the operation part designation information data having a very small data amount. Therefore, the amount of communication data between the client and server during parallel processing can be significantly reduced,
In addition, since the processing load on the client can be reduced, the parallel processing efficiency can be improved.

Although the amount of data to be transferred is the same as that of the conventional method, it is generally faster to transfer the data at a time when transferring the data at once than to transfer the data having the same amount of data into many parts. The purpose of the present invention is to improve the parallel processing efficiency by reducing the amount of communication data at the time of executing the parallel processing, and to shorten the entire processing time.

The first embodiment of the first invention of the present invention will be described. This embodiment is applied to a system for processing data divided into regions as shown in FIG. In FIG. 1, 11 is a client machine, 12-1, 12-2... Are server machines, 13 is a network, 14 is design data, 15 is drawing data, 16 is data of the whole chip, 16-1, 16-. Reference numeral 2 denotes each portion divided from the data 16 of the entire chip.

FIG. 2 shows a processing flowchart of the client 11. The client 11 expands the hierarchy of the LSI design data 14 input in step 101, performs processing of dividing the drawing into the meshes in the drawing unit area in step 102, and in step 103 divides the entire chip data into each server 12. -1, 12-2...

Next, in step 104, the position information is transferred to the server 12 for each divided area to be subjected to graphic processing. The server 12 extracts a figure in the corresponding area from the already transferred entire data according to the received position information,
The graphic processing is performed, and the resulting graphic data is transferred to the client.

The client 11 receives the graphic data from the server 12 (step 105) and constructs a drawing data file from the received data (step 10).
6). Steps 104 to 106 are repeated, and when the processing of all the divided areas is completed (Step 1)
07), the final drawing data file 15 is completed.

Next, a second embodiment of the first invention of the present invention will be described. This embodiment is applied to a processing system based on hierarchical processing as shown in FIG.
In FIG. 3, 11 is a client machine, 12-1,
12-2 ... is a server machine, 13 is a network,
Numeral 14 denotes design data, numeral 15 denotes drawing data, numeral 16 denotes data of the whole chip, and numerals 16-1 and 16-2 denote respective parts separated in cell units. A cell refers to a layout pattern having a set of functions when designing an LSI. Used as a standard in hierarchical design.

The processing flow of the client 11 is as shown in FIG. The client 11 performs processing such as removing overlap between cells in step 111 so that processing can be performed independently on each cell of the LSI design data, and transfers the processed data to each server 12. (Step 112).

Next, cell information to be processed by the server 12 is transferred to the server (step 113). Server 1
2 extracts a graphic in the corresponding cell from the entire data transferred in step 112 according to the information, performs a graphic operation process, and transfers the resulting graphic data to the client 11. The client 11 receives the graphic data from the server 12 (step 114) and constructs a drawing data file from the data received from the server (step 115). Steps 113 to 115 are repeated, and if the processing of all cells is completed (step 11
6), the final drawing data file is completed.

The second invention of the present invention addresses the second problem dealing with the addition or improvement of software functions as follows. That is, in the data processing in the second invention,
Specify the match condition and change condition, and specify the LS that matches the match condition.
By adding the function of selecting the I pattern data and changing the selected data as a change condition, it is possible to cope with the addition or improvement of the software function. Such a method eliminates the need to create software that is required only for the type of LSI pattern data processing, and enables desired LSI pattern data processing only by describing matching conditions and changing conditions in existing data. This is the aim of the present invention.

More specifically, 1 When software reads or outputs an LSI pattern data file such as GDSII STREAM, selection conditions and change conditions for figures, cells, cell references, and properties of individual transfer processes are specified. By setting, the data selected and changed according to the conditions are selected. 2. At this time, the selection condition and the change condition are represented by ASCII character strings, and the character strings are determined for each condition. By arranging them according to the determined rules, the conditions can be set and changed freely. . (3) By specifying a plurality of selection conditions, it is possible to specify a logical product and a logical sum condition. 4. Also, by specifying a plurality of change conditions, a plurality of data can be changed at one time. (5) In addition to executing such data change as a single unit, the data can be used by being incorporated in pre-processing and post-processing of software that inputs and outputs LSI pattern data.

As described above, a condition including the selection condition and the change condition is prepared as a filter-like function, and only by reading or outputting the LSI pattern data through the filter, the desired data can be selected or changed. Was executed simply and efficiently.

Here, the LSI pattern data and the LSI pattern data processing will be briefly described. LSI pattern data is data output from an LSI design CAD, and is generally GDSII STREAM FORMAT. It is composed of basic information units (hereinafter, referred to as elements) such as graphic information, cell citation information, and supplementary information, and a combination of the units forms a cell, and a combination of the cells forms an LSI pattern. Elements, for example, graphic information, are composed of basic data units (hereinafter referred to as records) such as graphic types, layer numbers, data type numbers, and coordinate values. The LSI pattern data finally follows the EB (El
ectoron Beam) data. Also,
In the LSI pattern data processing, for example, a boundary, which is one of the elements, is composed of a record of a figure type, a layer number, a data type number, and a coordinate value. The LSI pattern data processing is to change those values. In the present invention, data processing is performed on an element-by-element basis.
Alternatively, the processing is performed in units of records, and the processing is limited to processing that is possible in the process of reading records one by one and passing them to the next processing. The purpose of the change is to group patterns by semiconductor manufacturing process using the layer number, delete a pattern unnecessary for the semiconductor manufacturing process using the layer number, change the cell arrangement using the cell citation information, and the like.

Next, the matching condition and the changing condition of the present invention will be described. The matching condition is a matching condition with each element of the element and the record, and the matching condition is a processing target. The change condition is an instruction to change information held by the element and the record. The matching conditions are shown in FIG. 5 and FIG.
FIGS. 7 and 8 exemplify the change conditions.

Next, the basic operation of the matching condition and the changing condition will be described with some examples. Assuming that the start of the matching condition is ＠ and the start of the change condition is &, ＠ 0 & 1 indicates that the record of layer number 0 is changed to 1. The matching condition can be omitted, and if ＠ & 1, all records match the matching condition. On the other hand, if the change condition can be omitted and is set to マ ッ チ ン グ 0, nothing is processed, only matching is performed under the matching condition.

FIG. 9 shows a basic flowchart of such a basic operation. First, when the record unit data is read in step 201, it is checked whether or not this record matches the matching condition in step 202, and if it matches, the change condition is executed in step 203, and the record after execution of the change is read in step 204. replace. If it does not match the matching condition, it is left as it is.

In this case, basically, once the matching condition is passed, the data processing ends. However, according to the present invention, the matching condition and the changing condition are allowed to be repeated as follows, and when it is desired to create two elements of the layer number 1 and the layer number 2 from the element having the record of the layer number 0, You have to tell the matching condition to pass the element again. That is continue. $ 0 copy continue & 1 $ 0 & 2

FIG. 10 shows a flowchart in this case. First, when the record unit data is read in step 211, it is checked in step 212 whether or not the record matches the matching condition. If the record is matched, the change condition is executed in step 213, and the record after the change is executed is replaced. If it does not match the matching condition, the process proceeds to the next step.

Next, at step 214, the match condition is continued.
Is determined, and if it is included, the process returns to step 212 and this record is continued.
Checks whether the match condition after the is matched. Finally, in step 215, the record after the change is executed is replaced. If this function is used in reverse and {0} 1 & delete, elements other than layer numbers 0 and 1 will be deleted.

Next, an example of a logical product condition in the matching condition will be described. In this case, the conditions need only be separated by white space. For example, the following example indicates that BOUNDARY having a record of layer number 0 is changed to layer number 1. ＠ 0 boundary & 1

Next, an example of a logical sum condition in the matching condition will be described. Instead of the space in the previous logical product, you only need to separate them with or. For example, the following example is BOUNDAR
This indicates that the layer numbers of Y and PATH are changed to 1. ＠Boundary or path & 1

Next, an example in which a conditional statement is specified on the command line will be described. For example, when there is a command “convert” that only interprets and executes the matching condition and the changing condition, the input LSI pattern data file and the conditional statement are described on the command line, and the LSI pattern data processing is performed. This flowchart is shown in FIG. The condition execution in FIG. 11 is an execution flow of the matching condition and the change condition described in FIG. 9 and FIG.

In FIG. 11, the LSI layout data 14 is input in units of records in step 221, and
According to the conditions on the command line input from the console or the like in step 4, condition execution is performed in step 222.
After the execution is completed, the result is output on a record basis and input to the LSI layout data 14. Step 222 in FIG.
Is the execution flow of the matching condition and the change condition described in FIG. 9 and FIG.

Next, an example in which a conditional statement is specified by a file will be described. Write the conditional statements mentioned in the previous example in a file and specify them at the time of execution, or by default, for example, cond.txt
To be read. Reuse and sharing are easier. This flowchart is shown in FIG.

Next, an example in which a conditional statement is used as an input filter of a certain program will be described. For example, FIG.
As shown in FIG. 3, only a pattern having a necessary layer number is input before the graphic calculation program, and the graphic calculation can be performed only on the pattern. Thus, it can be used as an input filter for a certain program.

The third aspect of the present invention deals with the division of a graphic having repetition information as follows. First, when cutting a pattern pattern data group including repetition information, which is expressed as a file on a computer, in a region of interest, the original repetition information is saved as much as possible so that the subsequent processing can be performed. Try to avoid the burden.

Here, the specific contents of the third invention of the present invention will be described. First, there is a method for avoiding an increase in data due to clip processing. When a graphic having repetition information is clipped, the inner, outer, and clipping targets are not determined for all the developed graphics, and the repetition information is the area coordinates occupied by the unit repetition graphic, and the repetition pitch in the X and Y directions. In view of the relationship between the number of repetitions in the X direction and the Y direction and the clip region rectangle, repetition information included in the repetition region rectangle, repetition information outside (including those in contact with) the repetition region rectangle,
It is divided into repetition information that intersects with the repetition area rectangle (the above operation is called repetition clip), only the repetition information that intersects is developed, and more detailed clip processing for each figure is performed. By doing so, it is not necessary to develop and evaluate the clip area even to the figure completely included in the repetition, so that the clip processing time can be saved.
In addition, a portion that is rubbed on the outer peripheral portion can be represented by a maximum of four pieces of repetition information and is not expanded, so that the output data amount can be saved.

Next, a method of clipping repetition information will be described. Here, the details of the above-described repeated clip processing will be described using generalized mathematical expressions. Information on the repetition and the clip area is represented as follows. Clip area rectangular coordinates (C _x1 , C _y1 , C _x2 , C _y2 ) Repetition unit (lower left corner) Graphic area (a _x1 , a _y1 , a _x2 ,
a _y2) repeating number n _x, n _y repeat pitch d _x, as a result of repeated d _y graphic region,

[0055]

(Equation 1)

When the index position of the repetition relating to the clip is calculated from these coordinate values,

[0057]

(Equation 2)

The numerical value obtained here is the index within the repetition of the graphic intersecting the clip area rectangle. i
Is the outermost index area that intersects the clip area rectangle, and j is the innermost index area that intersects the clip area rectangle. This example is shown in FIG.
Shown in In the case of normal repetition in which the repetition pitch is equal to or greater than the width of the repetition unit graphic, the index that intersects the repetition rectangle is at most one, so it is not necessary to consider the two numerical values of i and j. As in the example
If the repetition pitch is smaller than the repetition figure width, there is a possibility of intersecting with a plurality of indices, so this is handled by two numerical values.

As a result of this calculation, if i _x2 or i _y2 is less than 0, or if i _x1 is less than or equal to n _x , or if i _y1 is less than or equal to n _y , all of the repetitive figures are clip region rectangles. And no need for clipping. Also, i _x1 and i _y1 are less than 0 and i _x2 is n _x
When i _y2 is equal to or _larger than n _y , all of the repetitive figures are inside the clip area rectangle, and there is no need for clip processing. Furthermore, i _x1 and i _x2 +1 are equal and
If i _y1 and i _y2 +1 are equal, the clip region rectangle is
Again, there is no need for clipping since it is in the repeating gap.

In other cases, it is highly probable that the repetitive figure intersects the clip area rectangle (it does not intersect if the sides of the clip area rectangle pass through the repetition gap), and clip processing is required. .

In clipping, the portion corresponding to the index outside the index (i _x1 , i _y1 , i _x2 , i _y2 ) is converted into a new repetition expression (up to four due to the tongue shape). _{Alternatively} , in the case of a clip that leaves the inside, a portion corresponding to an index inside (j _x1 , j _y1 , j _x2 , j _y2 ) is converted into a new repeated expression. These two indexes and this two
An index corresponding to a donut-shaped area that fits between two indices intersects the clip area rectangle, and thus is subject to clip processing of the graphic itself.

Here, a specific example will be described. First, the clip target is EB (Electoron Beam) drawing data (M
(Eg, EBES format), where the figure has repeated information directly. In this case, the above-described repetitive clipping method is directly applied, and a portion that does not intersect the clip region rectangle is output as a graphic with repetition information. All of the graphics that intersect the clip region rectangle are cut out and output in the clip region rectangle.

Next, a case where the clipping target is a set of figures (including a cell reference), such as LSI layout data (STREAM format, etc.), and the cell reference has repetitive information. (A) By applying the above-described repetitive clip method, a portion that does not intersect the clip region rectangle is output as a cell citation with repetition information. The following processing is performed for all cell citations that intersect the clip region rectangle. For all the figure sets that are the definition contents of the quoted cells, for the result of moving rotation etc. based on the cell citation information,
The inside / outside / intersection determination is performed, and classification is made into those to be deleted, those to be output as they are, and those to be clipped. If the data is a clip, and if the data is a graphic, the graphic is handed over to a clip area rectangle cutout operation. In the case of a cell quotation, since it may have repetitive information, the work after (A) is applied recursively to the cell quotation.

[0064]

As described above, according to the first or fourth aspect of the present invention, the entire data is transferred to the server side in advance in the client / server system, so that the client can transfer the data from the server to the server. There is no need to transfer graphic data, so that the network transfer load during parallel processing can be reduced and the processing speed can be increased. In addition, the processing load on the client is reduced, and the processing command delay to the server is reduced, so that the processing waiting time on the server is reduced, and the parallel processing efficiency can be improved.

According to the second or fifth aspect of the present invention, since the basic information unit that matches the selection condition is changed according to the change condition, only the type of LSI pattern data processing is required. Is no longer required, and desired LSI pattern data processing can be performed only by describing the matching condition and the changing condition. As a result, the man-hours for creating and managing software can be reduced to the conventional (1 / number of software, about 1/20). In addition, there is no need to use a CAD tool or a verification tool to process LSI pattern data.
/ 10.

According to the third or sixth aspect of the present invention, the original repetition information is preserved as much as possible when clipping the pattern data including the repetition information. , Processing can be performed, and an increase in clip processing time and an increase in output data size can be avoided. Further, by avoiding an increase in output data size, processing time in a post-process of clip processing can be saved. it can.

[Brief description of the drawings]

FIG. 1 is a block diagram of a client-server system according to an embodiment of the present invention.

FIG. 2 is a processing flowchart of the embodiment shown in FIG. 1;

FIG. 3 is a block diagram of a client-server system according to another embodiment of the present invention.

FIG. 4 is a processing flowchart of the embodiment shown in FIG. 3;

FIG. 5 is a table showing matching conditions in the software change processing of the present invention.

FIG. 6 is a table showing matching conditions in the software change processing of the present invention.

FIG. 7 is a table showing change conditions in software change processing of the present invention.

FIG. 8 is a table showing change conditions in software change processing according to the present invention.

FIG. 9 is a flowchart of an embodiment of a software change process according to the present invention.

FIG. 10 is a flowchart of another embodiment of the software change processing of the present invention.

FIG. 11 is a flowchart of one embodiment of pattern data processing according to the present invention.

FIG. 12 is a flowchart of another embodiment of pattern data processing according to the present invention.

FIG. 13 is a flowchart of another embodiment of the pattern data processing according to the present invention.

FIG. 14 is an explanatory diagram showing indices related to a repetition area and a clip area according to the present invention.

FIG. 15 is a block diagram of a conventional client-server system.

FIG. 16 is a processing flowchart of the example shown in FIG. 15;

FIG. 17 is an explanatory diagram showing a clip of a pattern having repetition information.

[Explanation of symbols]

11 client machine, 12 server machine, 1
3 network, 14 design data, 15 drawing data, 16 entire chip data, 16-1, 16-2
... Divided data of each part.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI H01L 21/30 502G 541C

Claims (6)

[Claims] An image processing apparatus comprising a plurality of servers of a client / server model including a plurality of workstations or personal computers connected by a network and having a parallel processing function by a plurality of servers, and drawing pattern data to generate drawing data. In the data processing system, the entire pattern data is transferred to the plurality of servers in advance, and the client transfers the designation information of the portion to be processed by the server, and the server transfers the graphics of the portion designated by the designation information. Data is extracted from the entire pattern data that has already been transferred and subjected to graphic operation processing and the like,
The drawing data processing system transfers the result to the client, and the client merges the result to create a final drawing data file. 2. A drawing which has a parallel processing function by a plurality of servers of a client / server model including a plurality of workstations or personal computers connected by a network, and processes pattern data to create drawing data. In the data processing system, by specifying a selection condition and a change condition for a basic information unit in each data transfer process of data reading and data output by software, a basic information unit that matches the selection condition can be changed according to the change condition. A drawing data processing system comprising a selection changing means for changing according to the contents. 3. A drawing data processing system for generating drawing data by processing pattern data, wherein a pattern data group including repetition information, which is expressed as a file on a computer, is cut out and extracted in a region of interest. A drawing data processing system, wherein processing is performed to save the original repetition information as much as possible in an extraction area, and the repetition information is processed in repetition units. 4. A drawing which has a parallel processing function of a plurality of servers of a client / server model including a plurality of workstations or personal computers connected by a network and processes pattern data to create drawing data. In the data processing method, the entire pattern data is transferred in advance to the plurality of servers, and the client transfers designation information of a part to be processed by the server, and the server transfers the designation information of the part designated by the designation information. The graphic data is extracted from the transferred entire pattern data and subjected to graphic calculation processing and the like,
A drawing data processing method, wherein the result is transferred to the client, and the client merges the result to create a final drawing data file. 5. A drawing which has a parallel processing function of a plurality of servers of a client / server model including a plurality of workstations or personal computers connected by a network and processes pattern data to create drawing data. In the data processing method, in a process of reading data by software and transferring data output individually, by specifying a selection condition and a change condition for a basic information unit, a basic information unit matching the selection condition is changed to the change condition. A drawing data processing method characterized in that the drawing data is changed according to the content of the drawing data. 6. A drawing data processing method for generating drawing data by processing pattern data, wherein a pattern data group including repetition information, which is expressed as a file on a computer, is cut out and extracted in a region of interest. And processing the drawing information so as to save the original repeating information as much as possible in the extraction area, and processing the repeating information in units of repetition.