Non-Volatile Memory Programming Anomaly Detection Before Final Verify
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing non-volatile memory devices face challenges in detecting programming anomalies during iterative programming operations, leading to inefficiencies and reduced reliability, as current methods often require completion of the final programming stage before anomaly detection and recovery can be initiated.
Innovation Solution
An iterative model using a recurrent neural network (RNN) processes data from each programming stage to determine the likelihood of anomalies, enabling early detection and recovery by comparing probability values to a threshold, thus minimizing latency and improving anomaly detection capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional anomaly detection methods are used, then the programming operation can be completed, but anomaly detection is delayed until the final stage causing increased latency and reduced reliability
Solution Approach 1:
The patent applies preliminary action by performing anomaly detection during intermediate programming stages rather than waiting for the final stage. The system continuously monitors programming parameters (current, voltage, time) at each stage and uses machine learning models to predict anomalies before they manifest as final programming failures, enabling early detection and recovery actions.
Solution Approach 2:
The patent implements feedback by establishing a closed-loop monitoring system that continuously compares actual programming parameters against expected values and adjusts detection sensitivity dynamically. The system provides real-time feedback about programming health status, allowing for immediate corrective actions when anomalies are detected, thereby improving both reliability and reducing latency.
2Manufacturing precision
If iterative programming stages are performed, then programming precision is improved, but the complexity of monitoring and detecting anomalies increases
Solution Approach 1:
The patent applies segmentation by dividing the anomaly detection system into modular components: one for each programming stage, each monitoring specific parameters relevant to that stage. Each stage has dedicated monitoring logic that independently analyzes current, voltage, and time parameters, making the overall complex system manageable through modular design while maintaining high programming precision across multiple stages.
Solution Approach 2:
The patent utilizes parameter changes by dynamically adjusting monitoring parameters and detection thresholds based on the current programming stage. The system adapts its monitoring focus and sensitivity levels according to the specific requirements of each programming stage, optimizing detection capability without uniformly increasing complexity across all stages.
3Loss of time
If early anomaly detection is implemented, then recovery time is reduced, but additional processing resources and computational overhead are required
Solution Approach 1:
The patent applies partial action by implementing anomaly detection with selective intensity - performing comprehensive monitoring during critical programming stages where anomalies are most likely to occur, while using lighter monitoring during less critical stages. The system adjusts the level of computational resources allocated based on the specific programming stage and risk assessment, achieving early detection without uniformly excessive resource consumption.
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting computational resource allocation and monitoring intensity based on real-time programming conditions. The system modifies detection sensitivity, sampling frequency, and model complexity levels according to the current programming stage and detected anomalies, optimizing the balance between early detection capability and computational resource consumption.
Data Source
AI summary
An iterative programming operation having at least 1 to n-th programming stages may be performed to program a non-volatile memory device having memory cells connected through a word line. The n-th programming stage may apply an n-th program voltage to the word line and generate an n-th verification result indicating a number of the plurality of memory cells having a threshold voltage at least meeting a particular verification voltage at the n-th programming stage. An n-th model stage of an iterative model may be performed to utilize the n-th verification result and n-th historical data associated with at least an (n−1)-th programming stage to determine an n-th probability. A programming anomaly may be determined based on the n-th probability at least meeting an n-th threshold probability. In response to the programming anomaly, the iterative programming operation may be stopped prior to completing a final programming stage thereof.