Virtual Sub Lot Tracking for Semiconductor Wafer Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current split lot tracking methods in semiconductor manufacturing are time-consuming and complex due to the need for nearly one thousand queries to obtain split lot records, affecting in-line controls and analysis efficiency.
Innovation Solution
A data tracking method and system that creates virtual sub lots based on split historical data, determining the maximum split number and virtual sub split number for each process station, allowing for efficient copying of process data from original wafer lots to virtual sub lots, thereby reducing query time and improving tracking efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If recursive lookup of previous historical data lot table is used to track split lots, then complete split lot record queries are implemented, but nearly one thousand queries are required making the process time-consuming and complex
Solution Approach 1:
The patent pre-calculates and stores the complete split lot hierarchy path (Lot_ID, TERMINAL_Lot_ID, SPLIT_LEVEL) in the slave lot table during data entry. This preliminary action eliminates the need for recursive lookups during queries, reducing query time from nearly one thousand iterations to a single direct database query while maintaining complete tracking accuracy.
2Ease of operation
If denormalization with previous process station lot numbers is used to implement tracking, then tracking processing is implemented, but thousands of data are generated at each process station making it impossible for a single data to record all split lot numbers
Solution Approach 1:
The patent segments the split lot tracking data into hierarchical levels using the SPLIT_LEVEL field, where each level represents a generation of split lots. This segmentation allows the system to handle thousands of split lots by organizing them in structured groups rather than requiring a single flat data structure, thus managing complexity while maintaining ease of tracking operations.
Solution Approach 2:
The patent adds a new dimension (SPLIT_LEVEL) to the traditional lot tracking data structure. Instead of attempting to store all split lot numbers in a single field, the system uses multiple dimensions (Lot_ID, TERMINAL_Lot_ID, SPLIT_LEVEL) to represent the hierarchical relationship, transforming the data structure from one-dimensional to multi-dimensional for better scalability.
3Reliability
If traditional database design is used for lot tracking, then lot numbers and process stations are recorded, but in-line controls and analysis efficiency suffer due to time-consuming queries
Solution Approach 1:
The patent performs preliminary data preparation by pre-calculating and storing the complete hierarchical path information (Lot_ID, TERMINAL_Lot_ID, SPLIT_LEVEL) in the slave lot table during initial data entry. This preliminary action ensures data integrity is maintained while enabling rapid queries for in-line controls and analysis, thus improving productivity without sacrificing reliability.
Data Source
AI summary
A data tracking method applied in semiconductor manufacturing is provided. Split historical data of a wafer lot is retrieved and it is determined whether the end of lot data of the wafer lot is read. If not, a maximum split number required for a split processing applied to the wafer lot is determined. Next, a virtual sub split number required for the current process station is determined according to the maximum split number and a current split number. At least one virtual sub lot for the current process station is created according to the virtual sub split number. Process data of an original wafer lot corresponding to the virtual sub lot is copied to the virtual sub lot and virtual sub lots for the next process station is continuously created when creation is complete.