Parallel Post-Inspection Analysis for Semiconductor Wafer Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor inspection systems face inefficiencies in throughput due to the time-consuming nature of post-inspection tasks, which can vary significantly based on the number and type of defects detected, leading to inconsistent inspection run times.
Innovation Solution
Decoupling post-inspection tasks from inspection runs allows for the performance of these tasks concurrently with new inspection runs, utilizing task scheduling to streamline the process and reduce overall inspection time by separating data analysis from data acquisition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If post-inspection tasks are performed sequentially after each inspection run, then analysis completeness is improved, but inspection throughput deteriorates
Solution Approach 1:
The system performs post-inspection tasks in advance or in parallel by decoupling them from the inspection run sequence. Defect data from previous inspection runs is processed while subsequent inspection runs are正在进行, effectively performing analysis work ahead of when it would traditionally be needed, thus improving throughput without sacrificing analysis completeness
Solution Approach 2:
The patent introduces a temporal dimension to the inspection process by overlapping inspection runs with post-inspection tasks. Instead of a linear sequential process, the system creates a parallel processing architecture where defect data processing occurs in a separate temporal stream, allowing multiple operations to coexist without interfering with each other
2Loss of information
If post-inspection tasks are performed immediately after inspection runs, then data freshness is improved, but inspection run time variability worsens
Solution Approach 1:
The system segments the inspection process into distinct independent modules: inspection run execution and post-inspection task processing. By separating these functions, the inspection run time becomes consistent and predictable, while post-inspection tasks are processed asynchronously without affecting the timing or variability of inspection runs
Solution Approach 2:
The patent introduces an intermediary defect data structure that serves as a buffer between inspection runs and post-inspection tasks. This intermediary allows data to be passed between the two processes without requiring immediate processing, enabling inspection runs to proceed at a steady pace while post-inspection tasks are processed at their own optimal pace
Data Source
AI summary
An optical inspection tool can automatically perform analysis/operations after the tool has generated data identifying defects (e.g. a defect list) from an inspection run of an object such as a semiconductor wafer. The tool can decouple post-inspection tasks from performing inspection runs so that one or more post-inspection tasks are performed on defect data from a previous inspection run while another inspection run is in progress. This can significantly improve the throughput of the tool when multiple inspections are performed, since the inspection run time effectively is shortened to include only the time the tool is actually used to acquire defect data. One or more post-inspection tasks can be performed, including, but not limited to, merging inspection runs, removing duplicate defects, removing straight-line false alarms, and characterizing defects.