Hybrid hotspot detection
View Patent ↗Aspects of the invention relate to hybrid hotspot detection techniques. The hybrid hotspot detection techniques combine machine learning classification, pattern matching and process simulation. A machine learning model, along with false hotspots and false non-hotspots for pattern matching, is determined based on training patterns. The determined machine learning model is then used to classify patterns in a layout design into three categories: preliminary hotspots, preliminary non-hotspots and potential hotspots. Pattern matching is then employed to identify false positives and false negatives in the first two categories. Process simulation is employed to identify boundary hotspots in the last category.
1. A method of hybrid hotspot detection, executed by at least one processor of a computer, comprising:
receiving data of layout patterns of a layout design;
classifying the layout patterns as preliminary hotspots, preliminary non-hotspots and potential hotspots based on machine learning;
identifying false positives in the preliminary hotspots and false negatives in the preliminary non-hotspots based on pattern matching; and
identifying boundary hotspots in the potential hotspots based on process simulation.
2. The method recited in claim 1 , further comprising:
determining hotspot information for the layout design based on the preliminary hotspots, the preliminary non-hotspots, the false positives, the false negatives and the boundary hotspots; and
storing the hotspot information in a tangible processor-accessible medium, or displaying the hotspot information on a tangible medium.
3. The method recited in claim 1 , wherein the machine learning comprises a support vector machine model.
4. The method recited in claim 3 , wherein the support vector machine model is derived by a model calibration system.
5. The method recited in claim 1 , wherein the potential hotspots is determined by using a predetermined value that defines a separation between hotspots and non-hotspots in a feature space.
6. The method recited in claim 1 , wherein the pattern matching is based on false hotspots and false non-hotspots determined by a model calibration system.
7. The method recited in claim 1 , wherein the machine learning comprises encoding layout patterns based on pattern densities.
8. A hybrid hotspot detection system, comprising:
a machine learning classification unit configured to classify layout patterns as preliminary hotspots, preliminary non-hotspots and potential hotspots using a machine learning model;
a pattern matching unit configured to identify false positives in the preliminary hotspots and false negatives in the preliminary non-hotspots; and
a process simulation unit configured to identify boundary hotspots in the potential hotspots.
9. The hybrid hotspot detection system recited in claim 8 , further comprising:
a hotspot determination unit configured to determine hotspot information based on the preliminary hotspots, the preliminary non-hotspots, the false positives, the false negatives and the boundary hotspots.
10. The hybrid hotspot detection system recited in claim 8 , wherein the machine learning model is a support vector machine model.
11. The hybrid hotspot detection system recited in claim 8 , wherein the pattern matching unit employs false hotspots and false non-hotspots determined by a model calibration system for pattern matching.
12. The hybrid hotspot detection system recited in claim 8 , wherein the potential hotspots is determined by using a predetermined value that defines a separation between hotspots and non-hotspots in a feature space.
13. One or more non-transitory computer-readable media storing computer-executable instructions for causing one or more processors to perform a method, the method comprising:
receiving data of layout patterns of a layout design;
classifying the layout patterns as preliminary hotspots, preliminary non-hotspots and potential hotspots based on machine learning;
identifying false positives in the preliminary hotspots and false negatives in the preliminary non-hotspots based on pattern matching; and
identifying boundary hotspots in the potential hotspots based on process simulation.
14. The one or more non-transitory computer-readable media recited in claim 13 ,
wherein the method further comprises:
determining hotspot information for the layout design based on the preliminary hotspots, the preliminary non-hotspots, the false positives, the false negatives and the boundary hotspots; and
storing the hotspot information in a tangible processor-accessible medium, or displaying the hotspot information on a tangible medium.
15. The one or more non-transitory computer-readable media recited in claim 13 , wherein the machine learning comprises a support vector machine model.
16. The one or more non-transitory computer-readable media recited in claim 13 , wherein the potential hotspots is determined by using a predetermined value that defines a separation between hotspots and non-hotspots in a feature space.
17. The one or more non-transitory computer-readable media recited in claim 13 , wherein the pattern matching is based on false hotspots and false non-hotspots determined by a model calibration system.