Semiconductor Process Simulation Queueing for Parameter Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of semiconductor manufacturing processes leads to time-consuming and expensive simulations, particularly when optimizing parameters for desired device characteristics, as existing methods require manual editing of simulation settings and can result in redundant operations, making it difficult to efficiently predict and improve semiconductor device performance.
Innovation Solution
A method is proposed that optimizes semiconductor process simulations by receiving input data, ordering parameters based on their values, generating a job tree and parameter table, and executing simulations through a queue, which reduces redundant operations and improves resource allocation, allowing for more efficient prediction and optimization of semiconductor device characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simulation is carried out with various values of parameters to perform simulation under various conditions, then simulation accuracy and comprehensiveness are improved, but simulation time and cost increase
Solution Approach 1:
The patent applies preliminary action by generating a job tree structure before executing simulations. The job tree organizes all simulation tasks, identifies dependencies between them, and pre-calculates the execution sequence. This allows the simulation system to process jobs efficiently without redundant computations, reducing simulation time while maintaining accuracy across various parameter conditions.
Solution Approach 2:
The patent segments the simulation process into discrete jobs organized in a tree structure. Each node in the job tree represents a specific simulation task with defined parameters, allowing independent processing and optimization. This segmentation enables parallel execution of independent jobs and facilitates efficient resource allocation, reducing overall simulation time while comprehensive parameter coverage is maintained.
2Adaptability or versatility
If complex simulation processes are performed based on parameters having various values to handle increasingly complex semiconductor processes, then simulation completeness is improved, but simulation time and cost increase
Solution Approach 1:
The patent implements dynamics by making the simulation process adaptive through the job tree structure. The system dynamically determines the execution sequence based on job dependencies and parameter relationships. This dynamic approach allows the simulation to maintain completeness for complex semiconductor processes while optimizing execution efficiency by avoiding redundant computations and enabling parallel processing where applicable.
Solution Approach 2:
The patent applies parameter changes by systematically varying simulation parameters according to the job tree structure. The job tree organizes parameter combinations in a hierarchical manner, allowing the system to efficiently explore the parameter space required for simulation completeness. This structured parameter variation maintains comprehensive coverage while reducing redundant parameter combinations, thereby improving simulation efficiency.
3Measurement precision
If manual editing of simulation settings is performed to optimize parameters, then simulation accuracy is improved, but operation complexity and time increase
Solution Approach 1:
The patent applies self-service by implementing an automated job tree generation system that creates the simulation execution plan without manual intervention. The system automatically analyzes simulation requirements, generates the job tree structure, determines execution sequences, and manages parameter variations. This automation maintains simulation accuracy while significantly reducing operation complexity and the time required to set up simulations.
Solution Approach 2:
The patent replaces the manual mechanical process of editing simulation settings with an automated computational system. The job tree generation algorithm automatically processes simulation requirements and generates the execution plan, substituting manual operations with automated computational logic. This substitution maintains simulation accuracy while eliminating the complexity and time associated with manual editing.
4Productivity
If redundant operations are eliminated to reduce simulation time, then simulation efficiency is improved, but simulation setup complexity increases
Solution Approach 1:
The patent applies preliminary action by performing the complex setup work of identifying and eliminating redundant operations during the job tree generation phase. The system pre-analyzes all simulation jobs, determines dependencies, and structures the execution plan to inherently avoid redundant computations. This preliminary structuring reduces simulation execution time without adding complexity during the actual simulation runtime.
Solution Approach 2:
The patent introduces the job tree structure as an intermediary between simulation requirements and execution. The job tree acts as a mediator that automatically handles the complexity of identifying and eliminating redundant operations through its hierarchical structure and dependency analysis. This intermediary absorbs the setup complexity, allowing efficient simulation execution without requiring users to directly manage the complexity of redundancy elimination.
Data Source
AI summary
A simulation method of a semiconductor process includes receiving first input data including values of a plurality of parameters received from the outside, setting a simulation using the plurality of parameters, improving the simulation by ordering the plurality of parameters, and executing the simulation by processing a job stored in a queue, wherein the optimizing includes generating a table with respect to the plurality of parameters, ordering the plurality of parameters based on the first input data, generating a job tree based on the plurality of ordered parameters, reconstructing the table based on the job tree, and generating the queue by storing the job in the queue based on the reconstructed table.