Semiconductor Task Scheduling to Prevent Processing Area Deadlock
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deadlock situations in semiconductor processing apparatuses, such as wafer cleaning apparatuses, occur due to the acid-water binding rule, causing downtime and impacting yield and normal operation.
Innovation Solution
A processing task start method and device that generates and analyzes processing area relationship lists to identify and avoid deadlock situations by determining startable tasks based on predetermined rules, ensuring materials in different processing areas are sequenced correctly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple processing tasks are executed concurrently in a semiconductor processing apparatus, then productivity is improved, but deadlock situations occur due to the acid-water binding rule causing processing areas to be occupied incorrectly
Solution Approach 1:
The system performs preliminary analysis of processing area relationships before task execution. By generating and analyzing processing area relationship lists in advance, the system identifies potential deadlocks before they occur, allowing preventive scheduling decisions to be made without waiting for actual conflicts to manifest during task execution
Solution Approach 2:
The system implements a feedback mechanism where the processing area relationship analysis results are used to adjust task scheduling decisions. The scheduling unit receives feedback about potential deadlocks from the relationship analysis and modifies the task execution plan accordingly, creating a closed-loop control system that continuously prevents deadlocks while maximizing throughput
2Productivity
If task scheduling is optimized to maximize throughput, then productivity is improved, but the complexity of determining startable tasks increases due to processing area relationship constraints
Solution Approach 1:
The scheduling problem is segmented into distinct analytical components: generating processing area relationship lists, analyzing relationships between tasks and processing areas, identifying startable tasks, and making scheduling decisions. This segmentation transforms a complex monolithic scheduling problem into manageable modular steps, reducing overall system complexity while maintaining optimization capability
Solution Approach 2:
The processing area relationship list serves as an intermediary data structure that mediates between task definitions and scheduling decisions. Instead of directly analyzing complex task interdependencies, the system uses this intermediate representation to simplify the identification of startable tasks, acting as a buffer that decouples task management from scheduling logic
3Reliability
If processing area occupancy is strictly controlled to prevent deadlocks, then reliability is improved, but the number of executable tasks decreases reducing productivity
Solution Approach 1:
The system dynamically changes the parameter of task startability based on processing area relationship analysis. Instead of using fixed occupancy rules that either allow or block tasks absolutely, the system adjusts task execution parameters according to the analyzed relationships, enabling fine-grained control that prevents deadlocks while maximizing the number of concurrently executable tasks
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The present disclosure provides a processing task start method and device in a semiconductor processing apparatus. The method includes generating a processing area relationship list of a current processing task of the semiconductor processing apparatus; determining all to-be-started processing tasks and generating a processing area relationship list for each of the to-be-started processing tasks, based on the processing area relationship list of the current processing task and the processing area relationship list of each of the to-be-started processing task, determining whether all the to-be-started processing tasks include a startable processing task according to a first predetermined rule, when all the to-be-started processing tasks include a startable processing task, selecting a target startable processing task from all startable processing tasks according to a second predetermined rule, and starting the target startable processing task. Therefore, at any time, specified materials of the target startable processing task and the current processing task will not appear in the two same processing areas and in the opposite sequence to avoid deadlock and improve production efficiency.