NAND Flash Memory Error-Based Refresh Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
NAND flash memory devices face issues with data degradation due to aged deterioration and read-disturb errors, leading to unnecessary refresh operations that shorten the device's lifespan and increase power consumption.
Innovation Solution
A semiconductor storage device that monitors error counts in blocks and performs refresh operations only when the error count exceeds a threshold, optimizing the refresh interval based on actual data corruption state to minimize unnecessary operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If refresh operation is uniformly performed based on number of readout times, then data corruption is reduced, but device lifespan is shortened and power consumption increases
Solution Approach 1:
The patent changes the parameter basis for refresh operations from uniform readout count to actual error count. The controller monitors error counts in real-time and dynamically adjusts refresh timing based on actual data degradation levels, extending device lifespan by avoiding unnecessary refresh operations while maintaining data integrity.
Solution Approach 2:
The patent implements feedback by continuously monitoring error counts and using this information to control refresh operations. The controller receives feedback about actual data corruption states and adjusts refresh timing accordingly, creating a closed-loop system that optimizes both reliability and device lifespan.
2Reliability
If refresh operation is uniformly performed based on number of readout times, then data corruption is reduced, but power consumption increases
Solution Approach 1:
The patent changes the control parameter from uniform readout count to actual error count, enabling the system to perform refresh operations only when necessary. This reduces power consumption by eliminating unnecessary refresh operations while maintaining data integrity through error-based triggering.
Solution Approach 2:
The patent applies partial action by performing refresh operations only on blocks that actually require them, rather than uniformly refreshing all blocks. This selective approach reduces overall power consumption while maintaining reliability by focusing refresh efforts where data corruption actually occurs.
3Duration of action of moving object
If refresh operation is performed based on error count monitoring, then device lifespan is extended, but processing amount and power consumption increase
Solution Approach 1:
The patent segments the monitoring process by dividing the storage device into multiple blocks and monitoring error counts independently for each block. This segmentation allows the controller to process and manage error information in smaller, more manageable units, reducing overall processing complexity while extending device lifespan through targeted refresh operations.
4Duration of action of moving object
If refresh operation is performed based on error count monitoring, then device lifespan is extended, but power consumption increases
Solution Approach 1:
The patent changes the triggering parameter from uniform time intervals to actual error counts, enabling the system to extend device lifespan by performing refresh operations only when data corruption actually occurs. This parameter change reduces power consumption by eliminating unnecessary refresh operations.
Solution Approach 2:
The patent implements self-service by having the system automatically monitor its own error counts and trigger refresh operations based on actual degradation states. This self-monitoring and self-adjusting mechanism extends device lifespan while optimizing power consumption by performing refresh operations only when genuinely necessary.
Data Source
AI summary
A method of controlling a nonvolatile semiconductor memory includes checking a first group at a first interval period, the first group including a plurality of blocks, and when a first block in the first group satisfies a first condition, assigning the first block to a second group. The method includes checking, at a second interval period, an error count of data stored in the second group, and when a second block in the second group satisfies a second condition, moving data stored in the second block to an erased block in which stored data is erased among the plurality of blocks.