JPH06181258A - Manufacture of semiconductor device - Google Patents

Abstract

PURPOSE:To make it possible to easily synthesize layout data while maintaining original graphics and graphic data amount without paying attention to a data name of each graphic data. CONSTITUTION:An information file by group 18 is created by judging whether there are graphic data for the same data name in synthesis layout files 13 and 14. Data names and data planting names of the synthesis layout files 13 and 14 are changed based on the information file by group 18, a data name conversion file 19 and a synthesis rule file 20, and primary files 22 and 23 consisting of graphic data to be taken in only are prepared. And the graphic data of the primary files 22 and 23 are simply synthesized and a synthesis file 25 is created.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for manufacturing a semiconductor device (LSI), and more particularly to synthesizing layout data having a hierarchical structure stored in different files.

Due to recent advances in LSI technology, the scale of LSI has become extremely large. Therefore, the LSI
When the layout design is performed, the LSI is divided into a plurality of parts, for example, functional blocks. Then, different layout personnel perform layout design for each functional block, store the designed layout data in the design file, respectively, and finish the layout design.

Then, the layout data of different design files created by the respective layout personnel are combined to create a desired LSI layout library. On this occasion,
Although the graphic data having the same data name exist in each layout data designed by each layout engineer, the graphic data may have different graphic patterns. This is because the layout designers are distant from each other in layout design, and the layout techniques of the layout designers are not completely unified. Also, when designing the layout of the LSI, other LS
Increasing number of cases where a part of I is used as it is is another cause of different graphic patterns between graphic data having the same data name.

In the layout library of one LSI, in order to avoid confusion of processing, only one type of graphic data can be defined for one data name. Therefore, when the layout data of each design file are combined, only one figure data can be fetched for the same data name. Therefore, the original figure designed by each person in charge of layout may not be maintained after the layout data is combined.

Therefore, when synthesizing the layout data of each design file, a synthesizing method capable of maintaining the original figure in the layout data is required.

[0006]

2. Description of the Related Art Conventionally, due to the large scale of LSI,
In this layout design, different layout personnel perform layout design for each functional block. The layout data created by each layout manager is
A hierarchical structure is formed in order to reduce the number of graphic data, and each graphic data is given a planting name consisting of coordinate data of a reference point and a data name.

[0007] Then, the layout data of the design file created by each person in charge of the layout is combined to perform LS.
I Create a layout library for manufacturing. At this time, if there is a case where graphic data having the same data name but different graphic patterns exists in each layout data,
By using two methods, graphic data having the same data name but different graphic patterns do not exist.

One method is to expand the hierarchy of graphic data other than any one of the graphic data having the same data name but different graphic patterns to eliminate the graphic data. In the other method, the operator visually checks the graphic data in each layout data one by one, and changes the planting name of the graphic data having the same data name but different graphic patterns, that is, the coordinate data of the data name and the reference point. To do.

[0009]

However, in the method of expanding the hierarchy of graphic data to eliminate the graphic data, the graphic data amount increases. For this reason, there arise problems that the computer processing time in the layout verification increases and that there is no direct access device such as a magnetic disk device that can store all the graphic data for the layout verification.

In addition, in the method of changing the planting name while checking each figure data one by one, since it is a manual work by the operator, mistakes are likely to occur and
The number of man-hours is also very large, which causes a problem that the process cannot be actually performed. Further, in this case, if the layout data are combined while overlooking the graphic data having the same data name but different graphic patterns, the LSI may be manufactured with the layout graphic changed and a defective product may be produced. To be

The present invention has been made in order to solve the above problems, and layout data can be easily combined while keeping the original figure and the figure data amount without being aware of the data name of each figure data. The purpose is to be able to.

[0012]

In order to achieve the above object, the present invention divides a semiconductor device into a plurality of parts and provides layout data having a hierarchical structure in which graphic data is defined by a graphic pattern and a data name. In the method of manufacturing a semiconductor device, which is created for each of the portions and synthesizes each layout data, for each graphics data in each layout data, the graphics patterns of the graphics data having the same data name are compared, and the graphics pattern is When they do not match, the graphic data having the data name are classified into the data name change group.

Then, regarding the graphic data having the same data name classified into the data name change group, the data name of the graphic data other than one of the graphic data is changed to a data name different from the data names of the other graphic data. After this,
Each layout data is composed.

[0014]

According to the present invention, for each figure data in each layout data, figure data having the same data name but different figure patterns are classified into a data name change group. Then, for each of the graphic data having the same data name classified into the data name change group, the data name of any one of the graphic data other than the graphic data is changed to the data name different from the data name of the other graphic data, and each layout data is changed. Is synthesized. That is, since the graphic data is not expanded, the original graphic and the graphic data amount are maintained, and the layout data can be easily combined without considering the data name of each graphic data.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be described below with reference to the drawings. FIG. 3 is a schematic diagram of a layout data synthesizing apparatus for implementing the present invention. The layout data synthesizing device 1 is composed of a CAD (Computer Aided Design) device, and includes a CPU 2, a magnetic disk device 3, a display 4, a keyboard 5, a magnetic tape device 6 for data input / output, and the like.

When the layout data composition process is activated by operating the keyboard 5, the CPU 2 executes the layout data classification process shown in FIG. 1 and then executes the layout data composition process shown in FIG.

That is, in step 12, the layout data stored in the design data files 10 and 11 shown in FIG. 1 is copied or format-converted as it is, and the composition layout files 13 and 14 are stored in the magnetic disk device 3. To do.

The design data files 10 and 11 store layout data for each functional block of the semiconductor device designed by different persons in charge of layout. The layout data of each design data file 10 and 11 is hierarchized by a plurality of graphic data defined by graphic patterns and planting names. The planting name is the coordinate data of the reference point and the data name.

FIG. 4 shows an example of graphic data. Figure 4
The graphic data 30 shown in (a) has a reference point 31, and is composed of one graphic pattern 32, 34, and 35 and four graphic patterns 33, respectively. The graphic data 40 shown in FIG. 4B has a reference point 41 and is composed of one graphic pattern 32 and four graphic patterns 33.

FIG. 5 shows the definition of the graphic type. The figure pattern 50 in FIG. 5A has four coordinate points 51.
It is defined by a quadrangle region surrounded by a line segment that sequentially connects 54 to 54. In the graphic pattern 60 of FIG. 5B, a line segment connecting the four coordinate points 51 to 54 in order has a width, and is defined by a rectangular frame area.

Next, at step 15, the data name is sequentially read from one of the composition layout files 13 and 14, and it is determined whether or not there is graphic data having the same data name in the other file. If there is no graphic data having the same data name, the process proceeds to step 17. If there are graphic data with the same data name, the differences in the graphic patterns of those graphic data are compared in step 16, the groups shown in Table 1 below are grouped, and the process proceeds to step 17. For comparison of graphic patterns, subtraction in graphic logic operation is used. In addition,
The comparison of the graphic patterns may be performed simply by comparing the graphic types and the coordinate points shown in FIG. At this time, the start point, the end point, and the way around each vertex may or may not match.

[0022]

[Table 1]

In Table 1, it is not necessary to change the data name of the graphic data of group 1 when synthesizing the data. For the graphic data of group 2, the graphic data are commonly used when the data are combined. Further, the graphic data of group 5 becomes a data name change group, and the data name of one graphic data is changed at the time of data composition.

In step 17, the grouping information of the data names of the composition layout files 13 and 14 is stored in the grouping information file 18 of the magnetic disk device 3. Next, in step 21, the layout file for synthesis 1
A primary file 22 consisting of only the graphic data to be imported by changing the data name and the data planting name of 3, 14
23 is created and stored in the magnetic disk device 3. The primary files 22 and 23 are created based on the grouping information file 18 and the data name conversion file 19 and the synthesis rule file 20 stored in the magnetic disk device 3 in advance. The data name conversion file 19 is a file in which rules for changing the data name and the planting name of the graphic data are written.

The synthesizing rule file 20 is a file showing how to handle graphic data grouped into groups 3 and 4 of the grouping information file 18, and in this embodiment there are the following three processing methods.

(1) Since all the graphic patterns of the graphic data of the groups 3 and 4 do not match, a data name change group is formed, and the data name of one graphic data is changed and both graphic data are fetched.

(2) The graphic data of group 3 is fetched, but the graphic data of group 4 is not fetched. (3) Import graphic data of group 4
The graphic data of is not imported.

Then, in step 24, the primary file 2
2 and 23 graphic data are simply combined and combined file 2
5 is created and the composite file 25 is stored in the magnetic disk device 3. Then, the composite data of the composite file 25 is output to the magnetic tape device 6 and the process is ended.

As described above, in the present embodiment, for each figure data in the composition layout files 13 and 14, figure data having the same data name but different figure patterns are classified into data name change groups, and a grouping information file is created. 18 was created. Then, when synthesizing the data, one of the graphic data having the same data name classified into the data name change group is changed to a data name different from the data names of the other graphic data, and each layout data is synthesized. Therefore, it is not necessary to be aware of the data name of each figure data, and since the hierarchy of figure data is not expanded, the original figure and figure data amount can be maintained, and all figure data can be stored in the magnetic disk device 3 for layout verification. It can be stored, and an increase in computer processing time can be suppressed. Also, because the original figure can be maintained,
Prevents the generation of defective products due to deformation of layout figures,
SI productivity can be improved.

Further, in this embodiment, the graphic data of the group 3 of the grouping information file 18 is fetched, and the graphic data of the group 4 is not fetched, so that the data for the manufacturing data can be fetched. By importing the graphic data of the group 4 of the grouping information file 18 and not importing the graphic data of the group 3, data for skeleton data, that is, for verification can be acquired.

[0031]

As described in detail above, according to the present invention,
There is an excellent effect that the layout data can be easily combined while maintaining the original figure and the figure data amount without paying attention to the data name of each figure data.

[Brief description of drawings]

FIG. 1 is a flowchart illustrating a layout data classification process according to an embodiment.

FIG. 2 is a flowchart showing a layout data synthesizing process according to an embodiment.

FIG. 3 is a schematic diagram of a layout data synthesizing apparatus for carrying out an embodiment.

4A and 4B are explanatory views showing an example of graphic data.

5A and 5B are explanatory diagrams showing definition of graphic types.

[Explanation of symbols]

 2 CPU 3 magnetic disk device 10, 11 design data file 13, 14 composition layout file 18 grouping information file 19 data name conversion file 20 composition rule file

Claims (1)

[Claims] 1. A semiconductor device is divided into a plurality of parts, layout data having a hierarchical structure in which graphic data is defined by a graphic pattern and a data name is created for each of the parts, and each layout data is combined. In the method for manufacturing a semiconductor device described above, for each figure data in each layout data, the figure patterns of the figure data having the same data name are compared, and when the figure patterns do not match, the figure data having the data name are given as data names. Data names of graphic data other than one of the graphic data having the same data name classified into the data name change group are different from the data names of other graphic data. A semiconductor device in which each layout data is combined after being changed Manufacturing method.