Memory Controller Fail Bit Check Garbage Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory systems face challenges in maintaining data reliability due to weak low-temperature data retention characteristics, particularly in portable electronic devices where data retention deteriorates over time, leading to potential data loss and system instability.
Innovation Solution
A memory system and method that perform a fail bit check operation on data stored in semiconductor memories and initiate a garbage collection operation on memory blocks with a high number of fail bits, identifying and copying valid data to a target block while erasing the victim block, thereby maintaining data integrity and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is stored in semiconductor memories without regular fail bit checking, then storage capacity is maintained, but data reliability deteriorates due to weak low-temperature data retention characteristics
Solution Approach 1:
The patent performs fail bit check operations periodically on data in semiconductor memories before data corruption occurs, particularly addressing low-temperature data retention issues. By proactively checking for fail bits and identifying at-risk data, the system prevents reliability deterioration rather than reacting to failures after they occur.
Solution Approach 2:
The system implements a feedback mechanism where fail bit check results are analyzed and used to trigger garbage collection operations selectively. The controller monitors fail bit counts and initiates data migration for blocks exceeding thresholds, creating a closed-loop system that continuously maintains data reliability based on actual memory condition feedback.
2Reliability
If garbage collection is performed on all memory blocks, then data reliability is improved, but processing time and system performance deteriorate
Solution Approach 1:
Instead of uniformly applying garbage collection to all memory blocks, the patent selectively targets only those blocks containing data with high fail bit counts. The controller analyzes fail bit distribution across different memory blocks and applies garbage collection operations locally to specific victim blocks, thereby maintaining reliability where needed while avoiding unnecessary processing in healthy blocks.
Solution Approach 2:
The system performs fail bit checks on all data but initiates garbage collection only for blocks exceeding a predetermined fail bit threshold. This partial action approach ensures that resources are concentrated on problematic data while avoiding the time penalty of processing all memory blocks uniformly, achieving a balance between thoroughness and efficiency.
3Reliability
If fail bit check operations are performed regularly on all data, then data reliability is enhanced, but energy consumption increases
Solution Approach 1:
The patent divides memory space into multiple blocks and performs fail bit checks selectively on specific segments rather than uniformly across the entire memory. By segmenting the memory space and using fail bit thresholds to identify problematic segments, the system concentrates energy expenditure on areas where reliability issues are most likely to occur.
Solution Approach 2:
The system performs comprehensive fail bit checking but only triggers energy-intensive garbage collection operations when fail bit counts exceed predetermined thresholds. This partial action strategy ensures that full reliability checking is maintained while avoiding excessive energy consumption associated with unnecessary data migration operations.
Data Source
AI summary
The present technology relates to a memory system and a method of operating the same. The memory system includes a memory device including a plurality of semiconductor memories, and a controller configured to control the memory device to select a victim block based on a fail bit number of some data, among data that is programmed in each of the plurality of semiconductor memories, corresponding to a specific program state, and configured to perform a garbage collection operation on the selected victim blocks.