Flash Memory Retention Monitoring via Dedicated Blocks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for detecting wearout and retention in NAND flash memory devices are insufficient, leading to data corruption due to high error rates and limited data integrity, as they fail to accurately track the finite write/erase endurance and retention, especially with wear leveling techniques that enhance wearout clustering and aliasing effects.

Innovation Solution

A method involving writing monitor data to memory blocks, reading it after a period, and determining retention behavior to output results, which includes using non-wear-leveled blocks with known data patterns to characterize errors and measure retention and endurance, allowing for accurate detection of wearout before data corruption occurs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Duration of action of moving object

If wear leveling techniques are used to distribute write cycles evenly across memory blocks, then write/erase endurance is improved, but aliasing effects and wearout clustering occur that reduce measurement precision

Engineering Contradiction:
Improvewrite/erase enduranceVSAvoidretention tracking accuracy
Core Design Contradiction:
Duration of action of moving objectVSMeasurement precision

Solution Approach 1:

The memory device is segmented into two distinct sets: monitor data blocks and user data blocks. Monitor blocks are dedicated exclusively to tracking wearout and retention characteristics, while user blocks handle normal data operations. This segmentation prevents aliasing effects by separating measurement functions from operational functions, allowing precise retention tracking without interference from wear leveling operations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Monitor data blocks serve as an intermediary mechanism between the wear leveling controller and the actual retention measurement process. These dedicated blocks provide a controlled environment for measuring retention characteristics without being subject to the same wear leveling operations that cause aliasing effects in user blocks, thus enabling accurate retention tracking while maintaining endurance benefits.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If monitor data is written to blocks that have undergone multiple write cycles, then retention behavior can be accurately measured, but the blocks may have accumulated wear that affects reliability

Engineering Contradiction:
Improveretention behavior detectionVSAvoidblock reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

Monitor data is written to dedicated blocks that are specifically prepared and isolated from normal wear leveling operations. These blocks undergo controlled write cycles to establish known wear states before retention measurement begins. This preliminary action ensures that the blocks are in a known, stable state suitable for accurate retention measurement while preventing further wear that would compromise reliability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

Different quality characteristics are applied to different parts of the memory device. Monitor blocks are optimized for retention measurement with controlled write histories and protected from further wear leveling operations, while user blocks continue to undergo normal wear leveling. This local differentiation allows monitor blocks to maintain high reliability for measurement purposes while user blocks maintain operational flexibility.

Inventive Principle:
Principle #3Local quality

3Device complexity

If traditional methods are used to track retention, then device complexity is reduced, but data integrity deteriorates due to high error rates

Engineering Contradiction:
Improvetracking system complexityVSAvoiddata integrity
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

A simplified copy of the retention tracking function is implemented using dedicated monitor data blocks with known data patterns. Instead of using complex algorithms to infer retention from user data, the system creates a dedicated measurement copy that directly tracks retention characteristics. This copying approach reduces device complexity by using simple read-compare operations while significantly improving data integrity through accurate retention measurement.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS8572443B2System, method, and computer program product for determining a retention behavior for at least one block of a memory device having finite endurance and/or retention
Publication Date: 2013.10.29 GLOBALFOUNDRIES US INC
  • US8572443B2 patent drawing
  • US8572443B2 patent drawing
  • US8572443B2 patent drawing

AI summary

A method according to one embodiment includes writing monitor data to at least one block of a memory device having finite endurance and/or retention; reading the monitor data after a period of time; determining a retention behavior of the at least one block based on the reading; and outputting a result of the determining. A memory device according to one embodiment includes a plurality of memory blocks having finite endurance and/or retention, at least one of the blocks having monitor data written therein; and circuitry for addressing the blocks. A system according to one embodiment includes a memory device having finite endurance and/or retention, the memory device comprising: a plurality of memory blocks, at least one of the blocks having monitor data written therein, wherein the at least one block has been written to a plurality of times prior to writing the monitor data; and circuitry for addressing the blocks.