Flash Memory Model Parameter Verification via Source-Drain Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for modeling flash memories are inefficient, requiring extensive debugging and costly quality assurance tools, making it difficult for inexperienced engineers to locate and correct errors, thereby increasing research and development cycles and costs.
Innovation Solution
A method involving designing a test key with a source, drain, and gate, testing it to obtain data, extracting model parameters, verifying the reasonableness of physical characteristics based on source-drain voltage and drain current relationships, and iteratively adjusting parameters until the model passes quality assurance, thereby ensuring accurate and efficient modeling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the whole model is inspected after modeling is completed, then the model quality can be ensured, but the debugging time and complexity increase significantly
Solution Approach 1:
The patent introduces intermediate verification checkpoints during the modeling process (after parameter extraction, after initial simulation) to detect errors early before final model completion. This preliminary action prevents errors from propagating to the final stage, reducing subsequent debugging time while maintaining model quality.
Solution Approach 2:
The patent implements a feedback mechanism where simulation results are continuously compared against expected characteristics, and model parameters are adjusted based on this feedback. This iterative feedback loop ensures model quality is maintained throughout the process rather than relying solely on final inspection.
2Difficulty of detecting and measuring
If experienced engineers perform model debugging, then the problem location accuracy improves, but the modeling cost increases
Solution Approach 1:
The patent implements automated error detection and localization systems that enable inexperienced engineers to identify and correct model errors independently. The system automatically traces error sources through simulation data, eliminating the need for expensive expert intervention while maintaining high problem location accuracy.
Solution Approach 2:
The patent replaces the manual expert debugging process with an automated computational system that uses algorithms to detect and localize errors. This substitution eliminates the need for human experts in the debugging process, reducing costs while maintaining or improving detection accuracy through systematic automated analysis.
3Measurement precision
If model parameters are adjusted iteratively to pass quality assurance, then the model accuracy improves, but the research and development cycle lengthens
Solution Approach 1:
The patent performs preliminary parameter optimization and validation at intermediate stages of the modeling process, rather than waiting until the end. This early parameter tuning prevents the need for extensive iterative adjustments later, reducing the overall R&D cycle while maintaining model accuracy.
Solution Approach 2:
The patent implements continuous feedback loops during parameter adjustment where simulation results immediately inform parameter modifications. This real-time feedback accelerates the convergence to accurate model parameters compared to traditional batch processing approaches, reducing the number of iterative cycles required.
Data Source
AI summary
The present disclosure provides a method for modeling, including: S1): designing a test key having a source, a drain, and a gate, and testing the test key to obtain test data; S2): extracting a model parameter according to the test data; S3): verifying reasonableness of a physical characteristic of the model parameter based on a relationship between a source-drain voltage and a drain current, if the reasonableness passes the verification, a model file is established and the method proceeds to S4), if the reasonableness fails the verification, the method returns to S2) to adjust the model parameter, until the reasonableness passes the verification; S4): performing quality assurance on the model file, if the model file passes the quality assurance, the modeling is completed, if the model file fails the quality assurance, the method returns to S2) to adjust the model parameter until the model file passes the quality assurance.