JP2000106337A - Method for generating/holding database, and electron beam exposure system using the method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for generating/holding a database, wherein the consistency to the feature of pattern information using an electron beam exposure device is maintained, while reducing the size of database, and an electron beam exposure system using the database of reduced size. SOLUTION: This method for generating/holding a database which provides a graphic pattern information for each region as data, and an electron beam exposure system which uses it comprises (A) the pattern information for each graphic is separated from division information (1)-(9) wherein the graphic is divided by a border line in a region for generation and/or holding, and (B) a unit representing the graphic position coordinate and/or graphic size is made variable.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for creating and / or holding a database for creating and / or holding a database, and an electron beam exposure system using the method. SUMMARY OF THE INVENTION The present invention provides a technology for enabling data compression of a database for processing pattern information used in an electron beam exposure apparatus or the like, and an electron beam exposure system using the same. INDUSTRIAL APPLICABILITY The present invention can be used in various fields for creating and maintaining a database, and various fields for performing drawing using an electron beam exposure system. For example, when manufacturing a mask used for manufacturing a semiconductor device or the like, the mask is used. It can be suitably used when drawing a pattern or the like.

[0002]

2. Description of the Related Art In a conventional electron beam exposure apparatus,
For example, in the case of an electron beam exposure apparatus for manufacturing a semiconductor mask, a pattern is formed on the mask by irradiating the mask with EB data (electron beam drawing data) serving as pattern information and irradiating the mask with an electron beam. . The EB data is usually generated from layout data.

That is, for example, as shown in FIG.
Generate At this time, when generating the EB data, a graphic operation process is generally performed using an intermediate file of the EB data as an input / output. As an intermediate file of the EB data, a database 14 corresponding to the graphic information of the EB data is used in order to efficiently perform the graphic operation.

The EB data has a data structure as described below, for example, taking an example.

[0005]

The EB data has the above-described structure, for example, and generally has the following features.

1) A figure is divided for each area. For example, as described above, the figure is divided for each of the areas (1) to (n).

2) The position coordinates and the size of the figure are represented by integral multiples of the minimum unit. For example, as described above, the minimum unit of the pattern is defined in the header portion, and the graphic (rectangular,
The size of the trapezoid is represented by an integral multiple of the minimum unit.

In a conventional database, the above 1), 2)
The data structure is determined according to the characteristics of the graphic information of the EB data.

On the other hand, with the miniaturization of semiconductor design rules, the number of figures of EB data is increasing, and the size of the database is increasing. Due to this,
There is a problem that the used amount of the disk increases. Further, when the graphic operation is processed on a plurality of computers connected via a network, there is a problem that the transfer time becomes long when the database is transferred over the network.

Therefore, in order to reduce such problems, it is desired to develop a technique for reducing the size of the database while maintaining consistency with the above-mentioned features 1) and 2) of the graphic information of the EB data. ing.

Before describing the present invention, a data structure of a graphic in a database according to the prior art will be described.

As described above, a figure is divided into regions. That is, the figure is divided by the boundaries of the areas. If a figure exists only in a single area, the data of the figure need only be the data of the figure in the area. , The data of the figure corresponding to the divided number is required.

Normally, graphic information is represented by a basic graphic, and is generally described by a rectangle and a trapezoid. Hereinafter, the data of the rectangular pattern information and the data of the trapezoidal pattern information will be described.

Describing the rectangular data structure, the rectangular pattern has a position (posx, posy) in the area,
It is represented by four values of the width (width) and height (height) of the rectangle. In the example of FIG. 5, the position (posx, posy) of one vertex (the lower left vertex in the figure) is (0, 0),
A width (width) of 300 and a height (height) of 5
00 (FIG. 5A).

If one value is represented by 32 bits, the size of one rectangular data is 32 × 4 = 128 bits.

That is, it is as shown below. (Data structure) posx posy width height (32 bits 32 bits 32 bits 32 bits) Data size: 32 × 4 = 128 bits (Specific example, see FIG. 5A) posx = 0, posy = 0, width = 300, height = 500

Next, a trapezoidal data structure will be described. The trapezoidal pattern is located at the position (posx, pos
y), and represented by the positions (x1, x2, x3, x4) of the vertices of the trapezoid and the height (height) (see the example of FIG. 5B). Note that trapezoids include an X trapezoid having two sides parallel to the X coordinate and a Y trapezoid having two sides parallel to the Y coordinate. Here, the X trapezoid will be described.

As described above, one trapezoid is represented by seven values. Therefore, if one value is represented by 32 bits, the size of one trapezoidal data is 32 × 7 = 224 bits.

That is, it is as shown below. (Data structure) posx posy x1 x2 x3 x4 (32 bits 32 bits 32 bits 32 bits 32 bits 32 bits 32 bits height 32 bits) Data size: 32 × 7 = 224 bits (Specific example, see FIG. 5 (b)) posx = 0, pos = 0 0, x2 = 400, x3 = 400, x4 = 200, height = 500

Next, a figure divided into two or more areas will be described. As described above, if a graphic spans a plurality of regions, the graphic is divided at the boundary of the region, and data of the divided figures is required.

First, a rectangle will be described. When a rectangular pattern crosses between regions, it is divided by the boundary line of the region, and thus rectangular pattern information for the number of divided regions is necessary.

Assuming that nx × ny rectangles are divided by a boundary line, (32 bits × 4 × nx × ny) bits
Data size is required.

In the example of FIG. 6, the rectangle is divided into four as a result of straddling the boundary line (FIG. 6A). That is, four (nx = 2, ny = 2) rectangles are divided by the boundary line. At this time, for each rectangle,
Two values are required as data. Speaking of a rectangle at the lower left of the figure (referred to as rectangle [1]), posx [1], po
There are four values of sy [1], width [1], and height [1], and four values are similarly set as data for each of the rectangles [2], [3], and [4].

That is, data in the case where nx × ny rectangles are divided by a boundary line are as follows. posx [1], posy [1], width [1], height [1] posx [2], posy [2], width [2], height [2] posx [nx], posy [ny ], Width [nx], height [ny]

In the case of a trapezoid, if ny figures are divided by a boundary line, the required data size is:
(32 bits × 7 × ny) bits. In the example of FIG. 6B, three areas are covered (ny =
The trapezoid of 3) is divided by the boundary line.

In this case, the data when the ny trapezoids are divided by the boundary line are as follows. posx [1], posy [1], x1 1, x2 1, x3 1, x4 1, height1 posx [2], posy [2], x1 2, x2 2, x3 2, x4 2, height2, posx [nx], posy [ny], x1 ny, x2 ny, x3 ny, x4 ny, height ny

As described above, conventionally, the data structure is adapted to the feature of the EB data, which has graphic information for each region, so that the graphic straddling the boundary of the region is divided by the boundary line. For example, in the graphic pattern illustrated in FIG. 7, the elongated rectangle (1) on the left side of the figure is divided into five, the central rectangle (2) is divided into four, and the trapezoid (3) on the right side is divided into three. Data for each of the later figures was needed. Reference numeral (4) indicates the boundary line of the area, and particularly reference numeral (5) indicates the dividing line of the figure. When the graphic is divided by the boundary in this manner, the divided graphic information is held, and the amount of data is larger than that of having information as one connected graphic.

In the graphic calculation processing, it is usually necessary to connect the figures divided by the boundary line in order to recognize the outline of the figure not divided by the boundary line of the area. . For this reason, it is necessary to increase the processing capability required for the graphic operation processing.

Next, a problem based on the minimum unit of a graphic pattern in the prior art will be described. Conventionally, in EB data, a minimum unit is set to represent the position and length of a figure. The minimum unit is set small enough to maintain required accuracy. The position and length of the figure are expressed as an integral multiple of this minimum unit.

With the above expression, the accuracy of the minimum unit of the EB data is maintained. Only one such minimum unit is held. However, a figure requiring a minimum unit level of accuracy is a part, and the position and length of most figures are represented by a value of 10 ⁿ times (n is an integer of 1 or more) the minimum unit. Therefore, in the bit string representing the position and length values,
The 10 to the power of n is redundant.

For example, consider a case where the position and length values are represented by 32 bits when the minimum unit is set to 0.01.
If 10 times 0.1 is used as a unit, 2 ³ <10 <
2 ⁴ So, the value of the position or length can be expressed by 32-3 = 29bit. If the unit is 100, 100 / 0.01
Since it is = 10 ⁴ , the value can be expressed by 32- (3 × 4) = 20 bits.

Specifically, when the minimum unit is 0.01 (posx = 1000 μm, posy = 200 μm)
m, width = 300 μm, height = 500 μ
To represent the rectangle of m), it is as follows.

Posx = 100000 ← 100000
× 0.01 = 1000 μm posy = 20,000 ← 20,000 × 0.01 =
200 μm width = 30000 ← 30000 × 0.01 =
300 μm height = 50000 ← 50000 × 0.01 =
500 μm Each of the above data is represented by 32 bits.

On the other hand, the same as when the minimum unit is 100 (posx = 1000 μm, posy = 200 μm, w
The rectangular representation of (idth = 300 μm, height = 500 μm) is as follows.

Posx = 10 ← 10 × 100 = 1000 μm posy = 2 ← 2 × 100 = 200 μm width = 3 ← 3 × 100 = 300 μm height = 5 ← 5 × 100 = 500 μm In this case, data can be expressed in 20 bits. Becomes

The following shows five rectangles R1 to R shown in FIG. 7 when the minimum unit is 0.01.
5 is an example of a data structure representing No. 5; Here, 32 bits
Because there are five rectangles with four values of
4) The data length is 5 × 640 bits. If 32 bits are used to represent a unit, a total of 640 + 32
= 672 bits.

That is, the data structure is as follows. posx = 0, posy = 0, width = 1000
0, height = 10000 posx = 20250, pos = 0, width = 1
0000, height = 10000 posx = 40500, pos = 0, width = 1
0000, height = 10000 posx = 60750, posy = 0, width = 1
0000, height = 10000 posx = 81000, pos = 0, width = 1
0000, height = 10000 The above posx, posy, width, height
Are represented by 32 bits.

Conventionally, the data size has been extremely large as described above.

[0040]

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to maintain consistency with the characteristics of pattern information such as used in an electron beam exposure apparatus. Creation of a database that can perform and reduce the size of the database
An object of the present invention is to provide a holding method and an electron beam exposure system using a database with a reduced database size.

[0041]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, a method for creating and holding a database according to the present invention is a method for creating and holding a database that gives pattern information of a figure as data for each area. The pattern information of a figure extending over two or more areas is also created and / or stored separately for each figure, and divided information obtained by dividing the figure by a boundary line of the area.

As a result, pattern information of a figure extending over two or more areas is not divided into figure data for each area, but is created and held by the data of the single figure and the division information. In this case, the data of the single figure is processed by the division information, and is given as pattern information of the figure for each area while maintaining consistency with the pattern information used in the electron beam exposure apparatus or the like. This makes it possible to compress the data size.

In addition, since a graphic which is not divided by a boundary line can be held in a database which is an input / output of the graphic operation processing, it is possible to eliminate, for example, a processing for input / output of the graphic operation processing. Can be simplified and speeded up.

Further, creation of another database of the present invention
The holding method is a method for creating and holding a database that provides pattern information of a figure as data, and is characterized in that a unit indicating a position coordinate and / or a size of the figure is made variable.

As a result, data that requires precision is represented by a precise unit, and a description of data that does not require a precise unit is represented by a rough unit.
This makes it possible to compress the data size.

The electron beam exposure system according to the present invention
A database relating to each of the above-described creation / holding methods is used. Thus, electron beam exposure can be performed using a database with a reduced data size.

The present invention can be preferably implemented in a form in which a database is created and / or held as basic pattern information such as a rectangle or a trapezoid. Also, based on the minimum unit of the length required for the precision of the pattern to be formed, the unit is variable so that the position coordinates and / or the size of each figure are expressed in units of an integral multiple of the minimum unit. It can be carried out in the following manner.

Japanese Patent Application Laid-Open No. 55-59720 discloses a technique for compressing drawing data of electron beam exposure by repeatedly using unit figures (or group of unit figures). Japanese Patent Application Laid-Open No. 5-226236 discloses a technique for setting a specific code independent of a trapezoidal pattern code for a frequently used oblique rectangle to reduce a data amount and a processing time. The gazette discloses a technique that divides a pattern into a plurality of blocks, codes which shape is included in the block to form a simple combination, and enables necessary data to be compressed. The present invention employs a method that is significantly different from the present invention.

[0049]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, the present invention will be further described by describing embodiments of the present invention. However, needless to say, the present invention is not limited to the embodiments described below.

Embodiment 1 Conventionally, information of a graphic extending over a plurality of areas is, for example,
The information is represented by information of nx × ny figures divided by the boundary of the area. In this embodiment, the present invention is applied, and the data is collected into a single figure as data.

Nx × ny divided by the boundary line of the area
An example will be described in which a number of rectangles are combined into one rectangle. The data structure when combined into one rectangle is as shown below.

(Data structure when rectangles straddling the boundary of the region are combined into one rectangle) posx [1] posy [1] nx ny width [1] width [2] ··· width [nx] height [1] height [2] · · height [ny]

Here, posx [1], posy [1]
Is data representing the positions of the rectangles that have been combined into one, and nx and ny are the number of areas (the number of divisions) spanned by the rectangle. width [1], width
[2],..., Width [nx] are the widths in each of the rectangular regions [1], [2],.
right [1], height [2], ..., hei
ght [ny] is the same for each area [1], [2],.
, [Ny].

If the above data is known, the rectangular information for each area can be processed and output. That is, from the combined rectangular information, this can be used as rectangular data for each area as in the related art. For example, in the case of the above example, if a rectangular data structure as shown below is read in the input of the graphic operation processing, it can be processed as one rectangle.

(Rectangular Data Structure at Input of Graphic Operation Processing) posx [1], posy [1], Σ _{i = 1} ^nx width
[I], Σ _{i = 1} ^ny height [i]

If the present invention is applied, based on the above, this can be converted into rectangular data for each area as in the conventional case. This method will be specifically described. That is, as a method of converting a rectangular data structure that is not divided by a region boundary line (a new rectangular data structure according to the present invention) into EB data rectangular data that is divided by a region boundary line, for example, the following method is used. A conversion algorithm can be used.

In the following description, setEBrect (re
gx, regy, posx, posy, width, h
(right) is a rectangle (posx, posy, width, he) in the area at the position (regx, regy).
(right). pitchx, p
“itchy” indicates the size of the region in the x and y directions.
posx and posy represent coordinates when regx and regy are used as reference origins.

(Algorithm for converting data combined into one rectangle into a rectangle of EB data for each area based on division information) regx = regx (x1y1) posx = posx [1] for (i = 1; i ≦ nx: i ++) ｛regy = regy (x1y1) posy = posy [1] for (j = 1; j ≦ ny: j ++) ｛setEBrect (regx, regy, posx, posy, width th [i], height] ) Regy = regy + pitchy posture = posy + height [j] -regy regx = regx + pitchx posx = posx + width [i] -regx}

Represents a loop. That is,
Then, i is substituted as an initial value, and i is increased each time the loop is executed once. This loop is executed until i exceeds Nx. Also, ~ forms a loop.

The flow of the above algorithm is as follows. Please refer to FIG. 1 and FIG. With respect to the x direction, a rectangular position (posx [1] pos
y [1]) is obtained. X of the obtained area
The position in the direction is regx. Posx representing the position in the x direction within the area is represented by posx
[1] is substituted. Are sequentially performed with the value of i being 1 to nx. In the y direction, the position of the rectangle (posx [1] pos
y [1]) is obtained. Y of the obtained area
The position in the direction is defined as “regy”. Posy indicates the position in the y direction within the area.
[1] is substituted. Are sequentially performed with the value of j being 1 to ny. In the x direction and the y direction, the position of the region (regx, r
egy) and the position of the rectangle in the area (posx, p
osy) has been determined, so the i-th width (w
id [i]) and the j-th height (heig
ht [j]), the rectangular data in the area is set using the above-described function. In the y direction, the size of the region (pitchy) is added to the region representing the position of the region, and the position re of the next region is obtained.
gy. Then, the rectangular position p in the next area
Find osy. In the x direction, the size (pitchx) of the region is added to regx representing the position of the region, and the position re of the next region is obtained.
gx. Then, the rectangular position p in the next area
Find osx.

Since the operation for y is performed by the operation for each x, the processing is completed when the operation for all x is completed, and rectangular data in each area is obtained.

If the position and length values are expressed by 32 bits, the size of the rectangular data when the present invention is applied to collectively represent the rectangles is 32 + 32 + 32 + 32 + (32 bits × nx) + (3
2 bits × ny) = 32 × (4 + nx + ny) bits. In this equation, the first four 32s are each posx,
It corresponds to posy, nx, ny, and (32 bits × nx)
Corresponds to width, and (32 bits × ny) is hei
ght.

In a rectangular data structure according to the prior art adapted to the structure of the EB data, (32 bits × 4 × nx × n
Since y) bits, if nx ≧ 2 and ny ≧ 2 in the case of a rectangle, the data can be compressed.

Embodiment 2 Next, an example in which the present invention is applied to a case where a trapezoidal pattern extending over a plurality of regions is represented will be described.

A trapezoidal pattern over an area is represented as follows. xL and xR indicate the x coordinate on the left side and the x coordinate on the right side.

The data structure when ny X trapezoids divided by the boundary line of the area in the conventional method are combined into one X trapezoid is as follows. Where xL
[I] represents the X coordinate value on the left side, and xR [i] represents the X coordinate value on the right side.

As for the X trapezoid, a figure straddling a boundary line parallel to the X coordinate is to be put together. If the X trapezoid is grouped by a figure that straddles the boundary line parallel to the Y coordinate, trapezoids and rectangles will be mixed in the grouped figure. Here, the case of grouping by the same type of X trapezoid is shown. is there.

(Data structure when trapezoids straddling the boundary of the region are grouped into one trapezoid; trapezoid straddling the boundary parallel to the X coordinate for the X trapezoid) posx [1] posy [ 1] ny height [1] height [2] ··· height [ny] xL [1] xL [2] ·· xL [ny] xL [ny + 1] xR [1] xR [2] ·· xR [ny] xR [Ny + 1]

In the graphic operation processing, if the following X trapezoidal data structure is read, it can be processed as one X trapezoid. (Rectangular data structure at input of graphic operation processing) posx [1], posy [1], xL [1], xR
[1], xR [ny + 1], xL [ny + 1], Σ _{i = 1}
^ny height [i]

As a method of converting the data structure of the X trapezoid not divided by the boundary of the region (a new data structure of X trapezoid according to the present invention) into the data of the X trapezoid of the EB data divided by the boundary of the region. For example, the following conversion algorithm can be used.

In the following description, setEBxtrap (r
egx, regy, posx, posy, x1, x2
x3, x4, height) is (regx, reg)
In the area at the position y), an X trapezoid (posx, pos
y, x1, x2, x3, x4, height). regx (x1y1), regy (x
1y1) represents the position of the area including posx [1] and posy [1]. “pitchx” and “pitchy” indicate the size of the region in the x and y directions. posx, posy
Represents coordinates when regx and reg are used as reference origins.

(Algorithm for Converting Data Combined into One X Trapezoid to X Trapezoid of EB Data for Each Area Based on Division Information) regx = regx (x1y1) (1) Reggy = regy (x1y1) (2) for (i = 1; i ≦ ny: i ++) ｛(3) x1 = xL [i] −regx x2 = xR [i] −regx x3 = xR [i + 1] −regx x4 = xL [i + 1] −regx posx = x1 (4) setEBxtrap (regx, regy, posx, posy, x1, x2, x3, x4, height [i]) (5) regy = regy + pitchy Posy = posy + height [i] -regy (6)｝

(3) to (6) form a loop.

The flow of the above algorithm is as follows. (1) With respect to the x direction, a rectangular position (posx [1] p
osy [1]). The position in the x direction of the obtained area is defined as regx. (2) With respect to the y direction, a rectangular position (posx [1] p
osy [1]). The position in the y direction of the obtained area is defined as “regy”. (3) Steps (4) to (6) are performed sequentially with the value of i being 1 to ny. (4) x1 to x4 representing the position of the trapezoid in the x direction in the area
Ask for. Then, po representing the position in the x direction within the region
The lower left x coordinate (x1) of the trapezoid is substituted for sx. (5) The position (reg) of the region in the x direction and the y direction
x, regy), the position of the rectangle in the area (posx, p
osy) and the x coordinate (x1 to x4) of the trapezoid are determined, so that the i-th height (height) in the x direction is obtained.
Using the above function together with the data of [i]),
X trapezoidal data in the area is set. (6) In the y direction, the size of the area, pitchy, is added to the region representing the position of the area, and the position re of the next area is obtained.
gy. Then, the position posy of the X trapezoid in the next area is obtained.

If the position and length values are represented by 32 bits, the size of the rectangular data when the present invention is applied and the X trapezoids are collectively expressed is 32 + 32 + 32 + 32 + (32 bits × ny) + (3
2 bits × (ny + 1) × 2) = 32 × (5 + nx ×
3) Bit. In this equation, the first three 32 are each posx,
It corresponds to posy and ny, and (32 bits × ny) is he
(32 bits × (ny + 1) × 2)
Corresponds to xL, xR.

In the X-trapezoidal data structure according to the prior art adapted to the EB data structure, (32 bits × 7 × ny)
Since ny ≧ 2 in the case of an X trapezoid, the data can be compressed.

Third Embodiment In the prior art, the unit for describing pattern data is only one minimum unit, but here, the present invention is applied to
The unit of the graphic pattern is divided into several stages. This is because the unit of length required to maintain accuracy differs for each figure.

Based on the size of each unit for each figure, the coordinate values and the length of the figure are expressed as integral multiples of the unit. Where the unit is large, it is possible to reduce the number of bits of an integer representing a coordinate or a length.

Here, assuming that the minimum unit is 0.01 μm,
An example in which a unit is set every ten times will be described. As a method of setting the unit, the unit may be set every 20 times or every 100 times according to the precision of the coordinates and the length of the pattern.

An example in which five rectangles R1 to R5 are arranged as shown in FIG. Here, an example of reducing the number of bits representing the coordinate value and length of a figure by adding a pattern unit to each figure will be described. An example in which the number of bits expressing the is reduced will be described later).

The number of bits that can be reduced per coordinate value and length value is as follows. When the unit is 0.01 μm, the coordinates and length can be represented by an integer of 32 bits. unit represents a unit. The number of bits is
Represents an integer bit number representing coordinates and length.

(Number of bits that can be reduced per one value of coordinate value and length) unit = 0.01... 32 bits [0 bit reduction / one value] unit = 0.1... 29 bits [3 bit reduction / One value] unit = 1... 26 bits [6 bits reduced / one value] unit = 10... 23 bits [9 bits reduced / one value] unit = 100... 20 bits [12 bits reduced / one value]

That is, for example, the unit is 1
If it is set to 00, the number of bits of the value of the coordinates and the length is 20 bi
t, compared to the case where the unit is 0.01,
This is a reduction of 12 bits.

Here, as shown in FIG. 7, when converting data in a case where five rectangles R1 to R5 are arranged, a pattern unit is added to each figure. A unit of unit = 100 is added to R1, a unit of unit = 0.1 is added to R2,
A unit of unit = 1 is added to R3, and unit =
A unit of 0.1 is added, and a unit of unit = 10 is added to R5. The number of bits of the data of each figure is as follows.

(R1) unit = 100 posx = 0, posy = 0, width = 1, hei
ght = 1 Since each figure is represented by 20 bits, the number of bits is 20 × 4 = 80.

(R2) unit = 0.1 posx = 2025, posy = 0, width = 10
00, height = 1000 Since each figure is represented by 29 bits, the number of bits is 29 × 4 = 116.

(R3) unit = 1 posx = 405, posy = 0, width = 10
0, height = 100 For this graphic, each is represented by 26 bits, so the number of bits is 26 × 4 = 104.

(R4) unit = 0.1 posx = 6075, posy = 0, width = 10
00, height = 1000 Since each figure is represented by 29 bits, the number of bits is 29 × 4 = 116.

(R5) unit = 10 posx = 81, posy = 0, width = 10, h
Eight = 10 Since each figure is represented by 23 bits, the number of bits is 23 × 4 = 92.

In this example, the total number of data bits is
There are five rectangles with four values, and bi for representing each figure
The t number is as described above, and the unit of the pattern (u
nit) is 32 bits, this 32 bi
Since t is required for each figure, the sum is as follows. (32 + 20 × 4) + (32 + 29 × 4) + (32 + 2
6 × 4) + (32 + 29 × 4) + (32 + 23 × 4) =
668 bits

Now, unit is 0.01, 0.1, 1,
The number of rectangles in each of 10, 100 is n00
1, n01, n1, n10, and n100. The total number of bits of the data representing the rectangle in this case is as follows. (32 + 32 × 4) × n001 + (32 + 29 × 4) ×
n01 + (32 + 26 × 4) × n1 + (32 + 23 ×
4) × n10 + (32 + 20 × 4) × n100

On the other hand, the total number of rectangular bits in the data structure according to the EB data is as follows. 32+ (32 × 4) × (n001 + n01 + n1 + n1
0 + n100) The first 32 of this expression is the number of bits representing the unit of the pattern, and (32 × 4) is rectang
le (x, y, width, height), that is, x,
This is the number of bits describing a function indicating four data of y, width, and height, and this is the same as the number of figures.

Therefore, the difference between the number of bits of data representing a rectangle in the new data structure and the number of bits of a rectangle in the data structure according to the EB data in this example is as follows. 32 + (− 32) × n001 + (− 20) × n01 +
(−8) × n1 + 4 × n10 + 16 × n100

As a result, the larger the number of figures whose pattern units are 10 or more, the smaller the overall data size. The number of bits (32bi) representing the unit added to each figure
t) may increase the overall data size,
This data structure is effective when there are many figures that do not require accuracy (pattern units need only be 10 or more).

Fourth Embodiment Also in this example, based on the size of each unit for each figure, coordinate values and lengths of the figure are expressed as integer multiples of the unit. Embodiment 3 is a case in which a unit of a pattern is added to each figure, and this embodiment is different from Embodiment 3 in that
This is an example in which figures are classified for each pattern unit, and the number of bits representing the coordinate value and length of the figure is reduced.

Also in this example, the minimum unit is 0.01 μm, and
An example will be described in which a unit is set every ten times. FIG.
As an example, a case where five rectangles R1 to R5 are arranged as shown in FIG. Here, figures are put together for each pattern unit.

Unit = 0.1 posx = 2025, posy = 0, width = 10
00, height = 1000 posx = 6075, posy = 0, width = 10
00, height = 1000 Here, each data is represented by 29 bits.

Unit = 1 posx = 405, posy = 0, width = 10
0, height = 100 Here, each data is represented by 26 bits.

Unit = 10 posx = 81, posy = 0, width = 10, h
Eight = 10 Here, each data is represented by 23 bits.

Unit = 100 posx = 0, posy = 0, width = 1, hei
ght = 1 Here, each data is represented by 20 bits.

The total number of data bits in this example is
Assuming that there are five rectangles having four values, and that 32 bits are used to represent the unit of the pattern, the following is obtained. (32 + 29 × 4 + 29 × 4) + (32 + 26 × 4) +
(32 + 23 × 4) + (32 + 20 × 4) = 636bi
t

Now, unit is 0.01, 0.1, 1,
The number of rectangles in each of 10, 100 is n00
1, n01, n1, n10, and n100. The total number of bits of the data representing the rectangle in this case is as follows. (32 + 32 × 4 × n001) + (32 + 29 × 4 × n)
01) + (32 + 26 × 4 × n1) + (32 + 29 × 4
× n10) + (32 + 23 × 4 × n100)

On the other hand, the total number of rectangular bits in the data structure according to the EB data is as follows. 32+ (32 × 4) × (n001 + n01 + n1 + n1
0 + n100) The first 32 of this expression is the number of bits representing the unit of the pattern, and (32 × 4) is rectang
le (x, y, width, height), that is, x,
This is the number of bits describing a function indicating four data of y, width, and height, and this is the same as the number of figures.

Therefore, the difference between the number of bits of data representing a rectangle in the new data structure and the number of bits of a rectangle in the data structure according to the EB data in this example is as follows. (12 × n01-32) + (24 × n1-32) + (3
6 × n10-32) + (48 × n100-32)

As a result, in the method of this embodiment, the minimum unit 1
The larger the number of figures of 0 times or more (pattern unit is 10 or more), the smaller the entire data size becomes.

[0106]

As described above, according to the present invention, the data size of an information processing database used in an electron beam exposure apparatus or the like can be reduced, and the area required for EB data and the like can be reduced from the compressed data. Data that gives pattern information of a figure can be restored for each image. Therefore, it is possible to maintain consistency with the features of the pattern information as used in an electron beam exposure apparatus. By compressing the size of the database, for example, it is possible to reduce the disk usage and to speed up network transfer. In the case of data for each region, it is necessary to perform a process of connecting at the boundary of the region when inputting the graphic operation process. You don't have to.

[Brief description of the drawings]

FIG. 1 is a flow chart for converting data into data (EB data) giving pattern information of a figure for each area in the first embodiment of the present invention.

FIG. 2 is a flow chart of converting data into data (EB data) that gives pattern information of a figure for each area in the first embodiment of the present invention.

FIG. 3 is a flow chart of converting data into data (EB data) giving pattern information of a figure for each area in the second embodiment of the present invention.

FIG. 4 is a diagram showing a basic graphic operation processing system.

FIG. 5 is a diagram illustrating an example of a data structure of pattern information of a graphic;

FIG. 6 is a diagram illustrating an example of a data structure of pattern information of a graphic straddling a boundary of an area.

FIG. 7 is a diagram illustrating an example of a figure divided for each area.

FIG. 8 is a diagram showing an example of a data structure representing five rectangles.

[Explanation of symbols]

11: layout data, 12: graphic operation, 1
3 ... EB data, 14 ... Database,
(1), (2), R1 to R5 ... rectangular, (3) ...
Trapezoid.

Claims (6)

[Claims] 1. A method of creating and holding a database for providing pattern information of a figure as data for each area, wherein the pattern information is divided into pattern information for each figure and division information obtained by dividing the figure by a boundary line of an area. A method for creating and / or holding a database characterized by creating and / or holding. 2. A database according to claim 1, wherein the pattern information of the figure is created and / or stored as a pattern information of a basic form.
Retention method. 3. A method for creating and holding a database for providing pattern information of a figure as data, wherein the unit for indicating the position coordinates of the figure and / or the size of the figure is variable. Retention method. 4. The method according to claim 1, wherein the position coordinates and / or the size of each figure are shown in units of an integral multiple of the minimum unit required for the precision of the pattern to be formed. The method for creating and maintaining a database according to claim 3. 5. An electron beam exposure system for performing electron beam exposure based on data of pattern information of a figure for each area, the pattern information for each figure and division information for dividing the figure by a boundary line of an area. An electron beam exposure system, wherein the electron beam exposure is performed using a database created and / or held separately. 6. An electron beam exposure system for performing an electron beam exposure based on data of pattern information of a figure, wherein a unit indicating a figure position coordinate and / or a figure size is made variable and / or held. An electron beam exposure system for performing the electron beam exposure according to a database created.