Flash Memory Write Location Selection Based on Health Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash storage systems face degradation due to repeated writing, leading to reduced storage capacity over time, as existing wear leveling techniques do not account for intrinsic health variations in flash memory locations caused by manufacturing process variations.
Innovation Solution
A method to select writing locations in flash memory based on a measure of health, which considers both the number of write pulses and intrinsic health, rather than just the number of writes, to extend the system's lifetime and reduce overhead storage requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If wear leveling tracks only the number of writes to each location, then implementation is simple, but it does not account for intrinsic health variations caused by manufacturing process variations
Solution Approach 1:
The patent changes the parameter used for wear leveling from simple write count to a composite health metric that includes write count, read count, and intrinsic health factors derived from manufacturing variations. This allows the system to account for locations that degrade at different rates due to manufacturing process variations, thereby improving reliability without excessive complexity
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring write and read operations, tracking health metrics for each location, and using this information to dynamically select optimal write locations. This feedback loop enables the system to adapt to actual degradation patterns and extend storage lifetime
2Reliability
If more comprehensive health metrics are used to select write locations, then storage lifetime is extended, but storage overhead increases
Solution Approach 1:
The patent applies partial action by not requiring complete characterization of every possible health factor. Instead, it uses a practical subset of metrics (write count, read count, and representative intrinsic health factors) that provide sufficient improvement in storage lifetime while keeping overhead manageable. This avoids the excessive action of tracking every possible degradation parameter
3Reliability
If write operations are optimized based on detailed health metrics, then charge leakage is reduced, but measurement and tracking complexity increases
Solution Approach 1:
The system uses self-service by leveraging existing controller resources and operational data to derive health metrics. Rather than requiring external measurement equipment or complex testing procedures, the controller utilizes information already available from normal read and write operations, along with simple intrinsic health indicators, to assess location health and optimize write operations for reduced charge leakage
Data Source
AI summary
For each of a plurality of locations in flash memory, a number of pulses required to change a value stored in that location is obtained. From the plurality of locations, a location to write to is selected using the obtained number of pulses. The selected location is written to.