Non-Volatile Memory Retention Control With Dynamic Activation Energy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory devices face challenges in accurately determining data retention time, leading to potential resource wastage or data loss due to underestimation or overestimation of data retention, which is exacerbated by aging and varying temperature conditions.
Innovation Solution
Incorporation of a dynamic activation energy table that accounts for different temperature values and device ages to improve the accuracy of data retention management, using a data retention management scheme integrated into the power on routine for non-volatile memory devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a fixed activation energy value is used for data retention calculation, then the calculation is simple and fast, but the accuracy of data retention time determination deteriorates due to aging and temperature variations
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed activation energy value to a dynamic activation energy table that adapts to changing conditions. The table is indexed by temperature and aging程度 (age), allowing the activation energy value to dynamically adjust based on current operating conditions, thereby maintaining high measurement precision without excessive complexity
Solution Approach 2:
The patent changes the parameter of activation energy from a constant value to a variable parameter that depends on temperature and aging. By storing multiple activation energy values in a table corresponding to different temperature and aging conditions, the system accurately reflects the changing physical characteristics of the memory device over time and across temperature ranges
2Reliability
If data retention management is performed frequently to prevent data loss, then data safety is improved, but resource wastage increases due to premature block recycling
Solution Approach 1:
The patent implements feedback by continuously monitoring the actual data retention characteristics of memory blocks and comparing them against the predicted values from the activation energy table. This feedback mechanism allows the system to adjust its data retention management strategies, recycling blocks only when truly necessary based on actual performance data, thereby preventing both data loss and resource wastage
Solution Approach 2:
The patent applies preliminary action by using the activation energy table to predict future data retention characteristics before they actually occur. This allows the system to proactively manage data retention by planning block recycling operations in advance, ensuring data safety while avoiding premature or unnecessary recycling that would waste resources
3Reliability
If block recycling is performed based on conservative retention estimates, then data loss is prevented, but productivity decreases due to unnecessary block recycling operations
Solution Approach 1:
The patent changes the parameter of retention time estimates from conservative fixed values to dynamic, condition-based predictions using the activation energy table. This allows the system to adjust recycling thresholds based on actual temperature and aging conditions, preventing data loss when needed while avoiding unnecessary recycling operations that would reduce productivity
Data Source
AI summary
A data retention management scheme is incorporated into the power on routine for non-volatile memory devices. Based on an activation energy, a data retention time is determined and used to decide on whether to perform a data retention operation, such as block recycling. To improve the accuracy of this process, a dynamic activation energy table that has different activation energy values for different temperature values and, for each temperature value, different activation energies for different device ages, such as beginning of life values and end of life values is used.