Semiconductor Pattern Modeling With ML for Shape Transformation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The semiconductor fabrication process faces challenges in accurately modeling and replicating patterns due to shape transformations during various sub-processes, such as photolithography and etching, which limits the precision of pattern formation and increases the complexity and cost of integrated circuit (IC) manufacturing.
Innovation Solution
A system and method utilizing machine learning (ML) models trained on pairs of design and physical pattern samples to estimate and refine pattern shapes, enabling precise modeling of semiconductor processes and reducing the time and cost associated with IC design and manufacturing by simulating both design and physical patterns with high accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional semiconductor fabrication processes are used to form patterns, then physical patterns can be manufactured, but the patterns have different shapes from designed patterns due to process variations and transformations
Solution Approach 1:
The system performs preliminary simulation of the semiconductor fabrication process using trained ML models before actual manufacturing. By predicting the physical pattern shape from the design pattern beforehand, the system can identify and correct shape deviations in advance, improving both manufacturing precision and reliability of pattern formation.
Solution Approach 2:
The system creates a virtual copy of the fabrication process through ML model simulation. The trained models replicate the transformation from design pattern to physical pattern, allowing virtual testing and optimization without physical trial-and-error, thereby enhancing pattern shape accuracy and consistency.
2Area of moving object
If the complexity of semiconductor fabrication process is increased to form patterns with reduced size, then integration density of IC can be increased, but the complexity of modeling and reproducing patterns accurately becomes more difficult
Solution Approach 1:
The system uses ML models to create virtual replicas of complex fabrication processes. By training models on process data, the system can simulate and analyze pattern transformations without physically executing every process step, reducing modeling complexity while maintaining accuracy for high-density pattern formation.
Solution Approach 2:
The system transforms complex process modeling into simplified image-to-image translation tasks. By converting process parameters into visual pattern transformations that ML models can process, the system reduces modeling complexity while accurately capturing the effects of process variations on pattern shapes.
3Measurement precision
If machine learning models are used to simulate design and physical patterns, then pattern formation precision can be improved, but the complexity of training and implementing ML models increases
Solution Approach 1:
The system uses ML models to create accurate virtual copies of design and physical patterns. By training models on paired pattern images, the system achieves high estimation accuracy while the models handle the complexity of process modeling internally, separating precision requirements from implementation complexity.
Data Source
AI summary
A system for modeling a semiconductor fabrication process includes at least one first processor and at least one second processor. The at least one first processor is configured to provide at least one machine learning (ML) model, which is trained by using a plurality of pairs of images of a design pattern sample and a physical pattern sample. The physical pattern sample is formed from the design pattern sample by using the semiconductor fabrication process. The at least one second processor is configured to provide an input image representing a shape of a design pattern and/or a physical pattern to the at least one first processor and to generate output data defining the physical pattern and/or the design pattern based on an output image received from the at least one first processor.