# US 2013/0159943 A1 · Machine learning approach to correct lithographic hot-spots

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2013/0159943 A1
- Document ID: US20130159943A1
- Status: Published application
- Kind: A1
- Application number: 13/328,994
- Filed: 2011-12-16
- Published: 2013-06-20
- Earliest claimed: 2011-12-16
- Assignees: International Business Machines Corporation
- Inventors: Kanak Behari Agarwal; Shayak Banerjee
- Web page: https://patentgrep.com/patent/US20130159943A1

## CPC

- G03F 7/70441: {Optical proximity correction [OPC]}
- G03F 7/70433: {Layout for increasing efficiency or for compensating imaging errors, e.g. layout of exposure fields for reducing focus errors; Use of mask features for increasing efficiency or for compensating imaging errors}
- G03F 1/36: Masks having proximity correction features; Preparation thereof, e.g. optical proximity correction [OPC] design processes
- G03F: PHOTOMECHANICAL PRODUCTION OF TEXTURED OR PATTERNED SURFACES, e.g. FOR PRINTING, FOR PROCESSING OF SEMICONDUCTOR DEVICES; MATERIALS THEREFOR; ORIGINALS THEREFOR; APPARATUS SPECIALLY ADAPTED THEREFOR; (phototypographic composing devices B41B; photosensitive materials or processes for photographic purposes G03C; electrophotography, sensitive layers or processes therefor G03G)

## Abstract

A method, system, and computer program product for machine learning approach for detecting and correcting lithographic hot-spots in an integrated circuit (IC) design are provided in the illustrative embodiments. A layout corresponding to the IC design is received at a machine learning model (ML model). At the ML model using a hardware component, a set of input objects is identified corresponding to a target shape in the layout. A retargeting value is predicted for the target shape using the set of input objects, such that applying the retargeting value to the target shape in the layout causes the target shape to be modified into a modified target shape, wherein printing the modified target shape instead of the target shape eliminates a lithographic hot-spot that would otherwise occur from printing the target shape in a printed circuit corresponding to the IC design.

## Claims

### Claim 1

A method for detecting and correcting lithographic hot-spots in an integrated circuit (IC) design, the method comprising:
receiving at a machine learning model (ML model) a layout corresponding to the IC design;
identifying, at the ML model using a hardware component, a set of input objects corresponding to a target shape in the layout; and
predicting a retargeting value for the target shape using the set of input objects, such that applying the retargeting value to the target shape in the layout causes the target shape to be modified into a modified target shape, wherein printing the modified target shape instead of the target shape eliminates a lithographic hot-spot that would otherwise occur from printing the target shape in a printed circuit corresponding to the IC design.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
training the ML model using a collection of training sets, a training set in the collection of training sets including a training layout of a training IC design, a training set of factors corresponding to the training layout, and a training set of retargeting values corresponding to the training set of factors in the training layout.

### Claim 3 (depends on 1)

The method of claim 1, wherein identifying the set of input objects comprises:
performing a shape processing operation on the target shape, the shape processing segmenting an edge of the target shape into a set of edge segments; and
calculating, for each edge segment in the set of edge segments, an input object in the set of input objects by computing a distance from another shape in a neighborhood of the edge segment in the layout.

### Claim 4 (depends on 3)

The method of claim 3, wherein the set of input objects corresponds to the edge segment, wherein the retargeting value modifies the edge segment.

### Claim 5 (depends on 4)

The method of claim 4, wherein a new set of input objects is identified for each edge segment in the set of edge segments, and wherein a new retargeting value is predicted for each edge segment using the corresponding new set of input objects.

### Claim 6 (depends on 1)

The method of claim 1, wherein the lithographic hot-spot is a lithographic anomaly at a location in the layout where a variation between a printed shape and the target shape exceeds a threshold.

### Claim 7 (depends on 1)

The method of claim 1, wherein the set of input object comprises a set of factors associated with the target shape in the layout, wherein the set of factors is determined based on a training set of factors used to train the ML model.

### Claim 8 (depends on 1)

The method of claim 1, wherein the ML model is an application, the hardware component is a processor, and the application executes using the processor.

### Claim 9 (depends on 1)

The method of claim 1, wherein the ML model is implemented to include the hardware component.

### Claim 10 (depends on 1)

The method of claim 1, wherein the ML model includes an artificial intelligence neural network.

### Claim 11

A computer usable program product comprising a computer usable storage device including computer usable code for detecting and correcting lithographic hot-spots in an integrated circuit (IC) design, the computer usable code comprising:
computer usable code for receiving at a machine learning model (ML model) a layout corresponding to the IC design;
computer usable code for identifying, at the ML model using a hardware component, a set of input objects corresponding to a target shape in the layout; and
computer usable code for predicting a retargeting value for the target shape using the set of input objects, such that applying the retargeting value to the target shape in the layout causes the target shape to be modified into a modified target shape, wherein printing the modified target shape instead of the target shape eliminates a lithographic hot-spot that would otherwise occur from printing the target shape in a printed circuit corresponding to the IC design.

### Claim 12 (depends on 11)

The computer usable program product of claim 11, further comprising:
computer usable code for training the ML model using a collection of training sets, a training set in the collection of training sets including a training layout of a training IC design, a training set of factors corresponding to the training layout, and a training set of retargeting values corresponding to the training set of factors in the training layout.

### Claim 13 (depends on 11)

The computer usable program product of claim 11, wherein identifying the set of input objects comprises:
computer usable code for performing a shape processing operation on the target shape, the shape processing segmenting an edge of the target shape into a set of edge segments; and
computer usable code for calculating, for each edge segment in the set of edge segments, an input object in the set of input objects by computing a distance from another shape in a neighborhood of the edge segment in the layout.

### Claim 14 (depends on 13)

The computer usable program product of claim 13, wherein the set of input objects corresponds to the edge segment, wherein the retargeting value modifies the edge segment.

### Claim 15 (depends on 14)

The computer usable program product of claim 14, wherein a new set of input objects is identified for each edge segment in the set of edge segments, and wherein a new retargeting value is predicted for each edge segment using the corresponding new set of input objects.

### Claim 16 (depends on 11)

The computer usable program product of claim 11, wherein the lithographic hot-spot is a lithographic anomaly at a location in the layout where a variation between a printed shape and the target shape exceeds a threshold.

### Claim 17 (depends on 11)

The computer usable program product of claim 11, wherein the set of input object comprises a set of factors associated with the target shape in the layout, wherein the set of factors is determined based on a training set of factors used to train the ML model.

### Claim 18 (depends on 11)

The computer usable program product of claim 11, wherein the computer usable code is stored in a computer readable storage medium in a data processing system, and wherein the computer usable code is transferred over a network from a remote data processing system.

### Claim 19 (depends on 11)

The computer usable program product of claim 11, wherein the computer usable code is stored in a computer readable storage medium in a server data processing system, and wherein the computer usable code is downloaded over a network to a remote data processing system for use in a computer readable storage medium associated with the remote data processing system.

### Claim 20

A data processing system for detecting and correcting lithographic hot-spots in an integrated circuit (IC) design, the data processing system comprising:
a storage device, wherein the storage device stores computer usable program code; and
a processor, wherein the processor executes the computer usable program code, and wherein the computer usable program code comprises:
computer usable code for receiving at a machine learning model (ML model) a layout corresponding to the IC design;
computer usable code for identifying, at the ML model using a hardware component, a set of input objects corresponding to a target shape in the layout; and
computer usable code for predicting a retargeting value for the target shape using the set of input objects, such that applying the retargeting value to the target shape in the layout causes the target shape to be modified into a modified target shape, wherein printing the modified target shape instead of the target shape eliminates a lithographic hot-spot that would otherwise occur from printing the target shape in a printed circuit corresponding to the IC design.