Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Enhanced Data Sampling and Feature Generation for Machine Learning-based Lithography Hotspot Detection
Technology scaling has increased the complexity of integrated circuit design. It has also led to more challenges in the field of Design for Manufacturing (DFM). One of these challenges is lithography hotspot detection. Hotspots (HS) are design patterns that negatively affect the output yield. Identi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
AUC Knowledge Fountain
2022
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | Technology scaling has increased the complexity of integrated circuit design. It has also led to more challenges in the field of Design for Manufacturing (DFM). One of these challenges is lithography hotspot detection. Hotspots (HS) are design patterns that negatively affect the output yield. Identifying these patterns early in the design phase is crucial for high yield fabrication. Machine Learning-based (ML) hotspot detection techniques are promising since they have shown superior results to other methods such as pattern matching. Training ML models is a challenging task due two main reasons. Firstly, industrial training designs contain millions of unique patterns. It is impractical to train models using this large number of patterns due to limited computational and memory resources. Secondly, the HS detection problem has an imbalanced nature; datasets typically have a limited number of HS and a large number of nonhotspots. This requires the use of data sampling techniques to choose the best representative dataset for training. In this thesis, we explore the problem of hotspot detection using machine learning. Specifically, we tackle the problem of data sampling where we introduce a method for dataset selection that enables the reduction of the training dataset size and improves the data balance between hotspot and non-hotspot patterns. In addition, we explore feature engineering using image gradients as a method of improving ML HS detection models. |
|---|