Cell-Aware Root Cause Deconvolution for Defect Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current root cause deconvolution methods in circuit defect diagnosis are limited by their reliance on scan diagnosis, which struggles to accurately identify cell internal defects, especially as semiconductor technology advances, leading to increased occurrences of cell internal defects and inefficient yield analysis.
Innovation Solution
A cell-aware yield analysis method that uses a processor to receive failure reports, perform logic diagnosis, determine a probability distribution of root causes based on layout information, and weight probabilities for cell internal suspects higher than interconnect suspects, enabling more accurate identification of root causes and guiding physical failure analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional root cause deconvolution based on scan diagnosis is used, then interconnect defect suspects can be analyzed, but cell internal defects are not accurately identified
Solution Approach 1:
The patent segments the defect analysis into two distinct categories: cell internal defects and interconnect defects. By separating the analysis of these two types, the system can apply specialized methods to each, improving the accuracy of cell internal defect identification while maintaining interconnect defect analysis capabilities.
Solution Approach 2:
The patent applies different probability weighting to different types of suspects based on their local characteristics. Cell internal suspects are weighted differently than interconnect suspects, allowing the system to adapt the analysis method to the specific nature of each defect type, thereby improving overall diagnosis accuracy.
2Reliability
If multiple defect suspects are reported for each failing device, then comprehensive coverage is achieved, but false positives increase and PFA efficiency decreases
Solution Approach 1:
The patent uses an iterative feedback mechanism where probability distributions are updated based on observed diagnosis reports. The system continuously refines its root cause probability estimates by comparing predicted suspect probabilities with actual diagnosis outcomes, thereby reducing false positives while maintaining comprehensive coverage.
Solution Approach 2:
The patent dynamically adjusts probability weights and parameters based on the type of suspect (cell internal vs. interconnect) and the observed diagnosis reports. By changing these parameters adaptively, the system optimizes the balance between comprehensive defect detection and minimizing false positives, thereby improving PFA efficiency.
3Ease of operation
If conventional diagnosis methods are used, then general defect locations can be identified, but specific root causes involving multiple physical layers cannot be resolved
Solution Approach 1:
The patent extends the analysis from a single-layer defect identification to a multi-dimensional approach that considers multiple physical layers simultaneously. By incorporating layout information and analyzing suspects across different layers, the system can identify specific root causes that involve bridging defects on several physical layers, thereby improving root cause identification accuracy.
Data Source
AI summary
Logic diagnosis is performed on failing reports of defective integrated circuits to derive a diagnosis report for each of the failing reports which comprise information of suspects. The suspects comprise cell internal suspects and interconnect suspects. A probability distribution of root causes for causing the defective integrated circuits is determined to maximize a likelihood of observing the diagnosis reports based on a probability for each of the suspects given each of the root causes and a probability for each of the diagnosis reports given each of the suspects. The probability for each of the diagnosis reports given each of the cell internal suspects is weighted higher than the probability for each of the diagnosis reports given each of the interconnect suspects.