IC Defect Diagnosis Using Dynamic Root Cause Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing defect diagnosis tools in integrated circuits often fail to timely identify and accurately predict systemic defects, relying on untimely physical failure analysis and hotspot analysis from physical inspection.
Innovation Solution
A computing system that utilizes electrical failure data to cluster features in a physical layout design, applying unsupervised machine learning to detect and confirm the physical root cause of defects through dynamic root cause detection, guiding physical failure analysis for confirmation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical failure analysis and hotspot analysis are used to identify defects, then defect location can be determined, but the diagnosis is untimely as it relies on physical inspection after manufacturing
Solution Approach 1:
The system performs preliminary defect diagnosis by clustering design features and electrical failures before physical manufacturing is complete. By analyzing the relationship between design features and electrical failures in advance, the system can predict potential defects and their locations without waiting for physical failure analysis, thus resolving the timing delay while maintaining diagnostic accuracy.
2Reliability
If traditional defect diagnosis tools are used, then manufacturing defects can be detected, but systemic defects cannot be timely predicted
Solution Approach 1:
The system establishes a feedback mechanism by clustering electrical failures with corresponding design features to identify patterns of systemic defects. This feedback loop enables the system to learn from detected failures and predict systemic defects across the manufacturing process, improving both reliability through pattern recognition and productivity through automated prediction without requiring physical inspection of each unit.
3Measurement precision
If physical inspection processes are performed to confirm defects, then accurate defect identification is achieved, but the process becomes complex and time-consuming
Solution Approach 1:
The system creates a virtual model by clustering design features with electrical failures to simulate and predict defect locations before physical inspection. This digital twin approach allows the system to identify high-probability defect locations that can then be targeted for physical inspection, reducing the need for comprehensive physical inspection while maintaining accurate defect identification through the clustered feature relationships.
Data Source
AI summary
This application discloses a computing system to identify suspected defects in a manufactured integrated circuit, which correspond to electrical failures detected by a test applied to the manufactured integrated circuit. The computing system can utilize the suspected defects in the manufactured integrated circuit to cluster features in a physical layout design describing the manufactured integrated circuit. Each cluster of the features corresponds to a candidate for a physical root cause of the suspected defects in the manufactured integrated circuit. The computing system can detect a physical root cause of the electrical failures in the manufactured integrated circuit based on the clusters of the features. A physical failure analysis process includes an inspection of the manufactured integrated circuit to confirm the physical root cause of the electrical failures in the manufactured integrated circuit corresponds to a systemic manufacturing fault in the manufactured integrated circuit.