Production Tool Yield Screening for Acceptability Checks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing procedure for determining the acceptability of newly-added production tools in semiconductor manufacturing lacks a unified standard and is heavily influenced by subjective factors, resulting in inaccurate and inefficient check results.
Innovation Solution
A standardized acceptability check method and system that involves comparing yield data from new and old production tools, using data analysis to categorize yield data into high or slightly higher yield categories, and eliminating these data points to improve accuracy and efficiency, with a fuzzy system model and K-Means clustering algorithm for data analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If yield data from new production tool is compared directly with old production tool using existing procedures, then tool acceptability can be determined, but the check results are affected by subjective factors and lack unified standards
Solution Approach 1:
The patent transforms the check procedure by changing the parameters used for comparison. Instead of directly comparing raw yield data, the system categorizes yield data into different categories (high yield category, slightly higher yield category, other categories) and applies different analysis methods to different categories. This parameter transformation approach standardizes the check process while maintaining accuracy by eliminating subjective factors through automated classification and comparison rules.
2Measurement precision
If all yield data points are included in the comparison, then comprehensive assessment is achieved, but the accuracy is reduced due to inclusion of high and slightly higher yield categories that mask actual performance differences
Solution Approach 1:
The patent extracts and separates high yield category and slightly higher yield category data points from the overall yield data set. By taking out these specific categories that tend to mask actual performance differences, the system enables more accurate comparison of the remaining data points. This extraction principle allows the check to focus on data points that truly reflect tool performance differences, improving measurement precision without requiring reduction of the overall data quantity.
3Productivity
If standardized check procedure with data categorization and elimination is implemented, then accuracy and efficiency are improved, but the check system becomes more complex
Solution Approach 1:
The patent segments the yield data analysis process into distinct stages: categorization stage (classifying data into high yield, slightly higher yield, and other categories), elimination stage (removing high and slightly higher yield categories), and comparison stage (analyzing remaining data points). This segmentation transforms a complex monolithic check procedure into manageable modular steps, improving efficiency by allowing parallel processing and automated decision-making at each stage while reducing the overall system complexity through clear separation of concerns.
Data Source
AI summary
The embodiments of the present application provide an acceptability check method and check system for newly-added production tools. The check method includes: performing, after obtaining several new tool yield data and several old tool yield data, data analysis on the several new tool yield data and the several old tool yield data, determining whether the several new tool yield data and the several old tool yield data belong to a high yield category or a slightly higher yield category, eliminating the corresponding new tool yield data and old tool yield data if “yes”, and taking the remaining new tool yield data and the remaining old tool yield data respectively as screened new tool yield data and screened old tool yield data; determining, based on the screened new tool yield data and the screened old tool yield data, whether the new production tool is acceptable.