Yield Excursion Root Cause Identification via Statistical Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying the cause of yield excursions in semiconductor manufacturing are expensive, time-consuming, and unsuitable for analyzing a small number of failing die, as they often require physical failure analysis and are not designed to identify a single dominant failing mechanism.
Innovation Solution
The use of statistical hypothesis testing to analyze logic diagnosis data alongside physical features in the design layout to efficiently identify the root cause of yield excursions, specifically by extracting and analyzing open features such as stacked vias, stress vias, and SWS open features, which allows for quicker and more accurate identification of the dominant defect mechanism.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical failure analysis is used to identify the cause of yield excursions, then measurement precision is improved, but loss of time and loss of energy increase
Solution Approach 1:
The patent uses logic diagnosis results as a copy or representation of the actual physical defects. Instead of directly performing physical failure analysis on failing dies, the system analyzes the diagnostic data which serves as a copy of the defect information, thereby reducing time and cost while maintaining identification accuracy
Solution Approach 2:
The patent replaces the mechanical/physical failure analysis process with a statistical analysis approach. Instead of physically examining failing dies through cross-sectioning and microscopy, the system uses statistical hypothesis testing on logic diagnosis data to identify defect causes, substituting a computational method for a physical one
2Measurement precision
If physical failure analysis is used to identify the cause of yield excursions, then measurement precision is improved, but loss of energy increases
Solution Approach 1:
The system uses logic diagnosis results as a copy of defect information, eliminating the need for energy-intensive physical examination processes while maintaining the ability to identify defect causes with high accuracy
Solution Approach 2:
The patent substitutes the energy-intensive physical failure analysis system with a low-energy statistical analysis system that processes digital data from logic diagnosis, dramatically reducing energy consumption while preserving measurement precision
3Measurement precision
If conventional analysis techniques are used to analyze large populations of failing die, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts and focuses on the most critical aspect of failure analysis - identifying the dominant failing mechanism - rather than attempting to analyze all possible defect types and rank multiple systematic yield limiters. This extraction approach simplifies the analysis process while maintaining precision for the primary defect cause
Solution Approach 2:
Instead of analyzing large populations to identify multiple yield limiters and ranking them, the patent inverts the approach by analyzing small samples of failing die to identify the single dominant cause. This inversion of the analysis strategy reduces complexity while achieving the same precision goal
4Loss of time
If a small number of failing die are selected for physical failure analysis, then loss of time is reduced, but measurement precision deteriorates
Solution Approach 1:
The system uses logic diagnosis results as a copy of defect information that can be analyzed from small samples. Since the diagnostic data captures the essential failure characteristics, analyzing a small number of failing die provides sufficient statistical power to identify the dominant defect cause with high precision
Solution Approach 2:
The patent changes the parameter being analyzed from physical defect characteristics (requiring PFA) to logic diagnosis results. This parameter change allows for precise identification of defect causes from small sample sizes, as the diagnostic data provides sufficient statistical information without requiring extensive physical analysis
Data Source
AI summary
Yield excursions in the manufacturing process today require an expensive, long and tedious physical failure analysis process to identify the root cause. Techniques are disclosed herein for efficiently identifying the root-cause of a manufacturing yield excursion by analyzing fail data collected from the production test environment. In particular, statistical hypothesis testing is used in a novel way to analyze logic diagnosis data along with information on physical features in the design layout and reliably identify the cause of the yield excursion.