A plurality of patches may be divided from the layout constituting the semiconductor chip. Each of the plurality of patches may have a plurality of hash IDs generated using a hash function for each of the aforementioned segments.

With miniaturized and complicated processes of the semiconductor chips, the number of patches to be performed the optical proximity correction increases. Therefore, in order to effectively reduce the TAT when performing the optical proximity correction, it may be desirable to reduce the number of patches used to perform the optical proximity correction. Further, in order to improve the accuracy of optical proximity correction, it may be desirable to increase the population of each segment in which the above-mentioned average value is calculated.

The optical proximity correction system and the operating method according to some example embodiments relating to this will be described in detail below.

FIG. 6 is a block diagram which shows the optical proximity correction system according to some example embodiments.

Referring to FIG. 6, the optical proximity correction system 100 may include a master device 120, slave devices 121 to 12n (here, n is a natural number), and patch blocks 131 to 13n. The master device 120 and the slave devices 121 to 12n may be implemented as separate computing devices, or the master device 120 and the slave devices 121 to 12n may each be implemented as a plurality of processor cores. Although not shown in the drawing, the optical proximity correction system 100 may further include a storage medium in which an optical proximity correction tool for executing the optical proximity correction is stored. The optical proximity correction system 100 may further include a memory into which the optical proximity correction tool is loaded when the optical proximity correction tool is executed.

In some example embodiments, the optical proximity correction tool that performs the optical proximity correction according to example embodiments may be a computer program product, e.g., a tangible product, that includes computer-readable program codes, or a computer program product, e.g., a tangible product, that includes non-transitory computer-usable media including computer-readable program codes.

The master device 120 may be connected to the plurality of slave devices 121 to 12n, to control the plurality of slave devices 121 to 12n. Each of the plurality of slave devices 121 to 12n may be connected to a patch block including the plurality of patches to control the plurality of patches.

In an example embodiment, the first slave device 121 is connected to the first patch block 131. The first patch block 131 includes a plurality of patches PA11 to PAn1. Further, an n^thslave device 12n is connected to an n^thpatch block 13n. The n^thpatch block 13n includes a plurality of patches PA1n to PAnn.

The plurality of patches connected to each of the plurality of slave devices 121 to 12n may be the smallest unit that may be processed by each of the plurality of slave devices 121 to 12n.

Herein, it will be understood that descriptions of some of the plurality of slave devices 121 to 12n may be similarly applied to the other remaining slave devices. Furthermore, it will be understood that descriptions of some of the plurality of patch blocks 131 to 13n may be similarly applied to the other remaining patch blocks.

FIG. 7 is an example diagram for explaining a segment information table.

Referring to FIGS. 6 and 7, each of the plurality of patches included in each of the plurality of patch blocks 131 to 13n may include segment information tables. Herein, although the first patch block 131 will be described as an example, it will be understood that the description thereof may also be applied to the remaining patch blocks.

Each of the plurality of patches PA11 to PAn1 included in the first patch block 131 includes segment information tables 141 to 14n.

For example, a first patch PA11 includes the segment information table 141, and an n^thpatch PAn1 includes a segment information table 14n.

The segment information table includes a hash ID generated using a hash function for a plurality of segments generated from each of the plurality of patches.

For example, the segment information table 141 in the first patch PA11 includes hash IDs such as 0x78, 0x81, 0x93, and 0x1A. Further, the segment information table 14n in the n^thpatch PAn1 includes hash IDs such as 0x78, 0x85, and 0x93.

The segment information table includes number information (# of hash ID), e.g., a count, that is obtained by counting each of the plurality of hash IDs included in each of the plurality of patches. For example, the number information (# of hash ID) may be a count of a number of occurrences of the hash ID in the patch.

For example, referring to the example in FIG. 7, when the result of counting the 0x78 hash ID existing in the first patch PA11 is counted as nine, the segment information table 141 in the first patch PA11 records the counting number 9 for the 0x78 hash ID. When the result of counting the 0x81 hash ID existing in the first patch PA11 is counted as five, the segment information table 141 in the first patch PA11 records the counting number 5 for the 0x81 hash ID. When the result of counting the 0x93 hash ID existing in the first patch PA11 is counted as two, the segment information table 141 in the first patch PA11 records the counting number 2 for the 0x93 hash ID. When the result of counting the 0x1A hash ID existing in the first patch PA11 is counted as two, the segment information table 141 in the first patch PA11 records the counting number 9 for the 0x1A hash ID.

Still referring to the example in FIG. 7, when the result of counting the 0x78 hash ID existing in the n^thpatch PAn1 is counted as eight, the segment information table 14n in the n^thpatch PAn1 records the counting number 8 for the 0x78 hash ID. When the result of counting the 0x85 hash ID existing in the n^thpatch PAn1 is counted as seven, the segment information table 14n in the n^thpatch PAn1 records the counting number 7 for the 0x85 hash ID. When the result of counting the 0x93 hash ID existing in the n^thpatch PAn1 is counted as three, the segment information table 14n in the n^thpatch PAn1 records the counting number 3 for the 0x93 hash ID.

Further, the segment information table includes the number (Total # of hash ID in patch) obtained by counting the total types of the plurality of hash IDs included in each of the plurality of patches.

For example, when the number of total types of hash IDs existing in the first patch PA11 is counted as a total of 2,000 including 0x78, 0x81, 0x93, and 0x1A, the segment information table 141 in the first patch PA11 records the number of total types of hash IDs existing in the first patch PA11 as 2,000 for each hash ID.

Further, e.g., when the number of total types of hash IDs existing in the n^thpatch PAn1 is counted as a total of 2,500 including 0x78, 0x85, and 0x93, the segment information table 14n in the n^thpatch PAn1 records the number of total types of hash IDs existing in the n^thpatch PAn1 as 2,500 for each hash ID.

FIG. 8 is a flowchart for explaining the above-described segment information table generation operation.

Referring to FIGS. 7 to 8, hash IDs, which are generated using a hash function for a plurality of segments generated from each of the plurality of patches, are recorded in the segment information table (S210).

After that, the number (# of hash ID) information obtained by counting each of the plurality of hash IDs included in each of the plurality of patches is recorded in the segment information table (S220).

After that, the number (Total # of hash ID in patch) obtained by counting the number of total types of the plurality of hash IDs included in each of the plurality of patches is recorded in the segment information table.

The order of steps S220 and S230 may be changed from each other or may be performed in parallel.

FIG. 9 is an example diagram for explaining a minimum patch table.

Referring to FIGS. 6, 7, and 9, each of the plurality of slave devices 121 to 12n may receive a plurality of segment information tables from patch blocks 131 to 13n connected to each of the plurality of slave devices 121 to 12n to generate the minimum patch table.

Herein, although the operation of the first slave device 121 and the n^thslave device 12n will be described, it will be understood that the description thereof may be applied to the remaining slave devices.

The first slave device 121 receives a plurality of segment information tables from the plurality of patches PA11 to PAn1. After that, the first slave device 121 generates the minimum patch table 151, using the plurality of segment information tables received from the plurality of patches PA11 to PAn1.

Herein, the segment information tables received by the first slave device 121 will be described as, e.g., the segment information tables 141 to 14n of FIG. 7.

The first slave device 121 compares the same hash IDs among the plurality of hash IDs on the plurality of segment information tables 141 to 14n received from the plurality of patches PA11 to PAn1, and records the patch having the larger total types of number (Total # of hash ID in patch) in the minimum patch table 151.

For example, the first slave device 121 compares the same 0x78 hash IDs in the segment information table 141 to the segment information table 14n. At this time, 2,500, which is the number of total types for 0x78 hash ID of the segment information table 14n in the n^thpatch PAn1, is assumed to be the highest. Therefore, the first slave device 121 records information, which is stored in the segment information table 14n of the n^thpatch PAn1, in the minimum patch table 151 as a 0x78 hash ID.

Similarly, the first slave device 121 compares the same 0x85 hash ID in the segment information table 141 to the segment information table 14n. At this time, 2,500, which is the number of total types for 0x85 hash ID of the segment information table 14n in the n^thpatch PAn1, is assumed to be the highest. Therefore, the first slave device 121 records the information, which is stored in the segment information table 14n of the n^thpatch PAn1, in the minimum patch table 151 as a 0x85 hash ID.

Similarly, the first slave device 121 compares the same 0x81 hash IDs in the segment information table 141 to the segment information table 14n. At this time, 2,000, which is the number of total types for 0x81 hash ID of the segment information table 141 in the first patch PA11, is assumed to be the highest. Therefore, the first slave device 121 records the information, which is stored in the segment information table 141 of the first patch PA11, in the minimum patch table 151 as a 0x81 hash ID.

Similarly, the first slave device 121 compares the same 0x93 hash ID in the segment information table 141 to the segment information table 14n. At this time, 2,500, which is the number of total types for 0x93 hash ID of the segment information table 14n in the n^thpatch PAn1, is assumed to be the highest. Therefore, the first slave device 121 records the information, which is stored in the segment information table 14n of the n^thpatch PAn1, in the minimum patch table 151 as a 0x93 hash ID.

Similarly, the first slave device 121 compares the same 0x1A hash IDs in the segment information table 141 to the segment information table 14n. At this time, 2,000, which is the number of total types for the 0x1A hash ID of the segment information table 141 in the first patch PA11, is assumed to be the highest. Therefore, the first slave device 121 records the information, which is stored in the segment information table 141 of the first patch PA11, in the minimum patch table 151 as a 0x1A hash ID.

A type in which the first slave device 121 records the minimum patch table 151 may also be applied to a type in which the n^thslave device 12n generates the minimum patch table 15n using a plurality of segment information tables received from the 1n^thpatch PA1n to the nn^thpatch PAnn. Therefore, the description of the minimum patch table 15n will not be provided.

FIG. 10 is a flowchart for explaining the above-described minimum patch table generation operation.

Referring to FIGS. 6, 9, and 10, each of the plurality of slave devices 121 to 12n compares the same hash IDs among the plurality of hash IDs, with reference to the segment information tables received from the plurality of patches (S310).

At this time, for the same hash IDs, a patch including a hash ID having a higher total type of hash ID is selected and recorded in the minimum patch table (S320).

Because each of the plurality of slave devices 121 to 12n selects only the patch having the higher total type of hash ID for the same hash ID with reference to the segment information table received from the plurality of patches, it is possible to reduce the runtime or TAT for the optical proximity correction system according to some example embodiments when performing the optical proximity correction operation.