Flash Storage Control for Marginal Block Tagging and Wear Balance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current data management systems, particularly those using flash memory, face challenges in maintaining reliability and efficiency as erase blocks degrade, leading to marginal blocks that are unusable due to failure conditions such as bit errors and uneven wear, which affects the overall performance and longevity of solid-state disks.

Innovation Solution

A storage control system that generates encoded data with a proportional data distribution for writing, identifies marginal blocks by reading erase blocks, and assigns a marginal tag with a non-proportional data distribution to mark unusable blocks, ensuring they are not used for normal operations and allowing for easier identification and maintenance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Volume of stationary object

If a large number of flash memories are provided to achieve the same storage capacity as HDD, then storage capacity is maintained, but data management complexity and reliability deterioration increase due to uneven wear and marginal blocks

Engineering Contradiction:
Improvestorage capacityVSAvoiddata management reliability
Core Design Contradiction:
Volume of stationary objectVSReliability

Solution Approach 1:

The system performs preliminary actions by generating encoded data with proportional data distribution before writing to flash memories, and by identifying marginal blocks before they cause failures. This proactive approach prevents reliability deterioration by balancing wear patterns and detecting degraded blocks early in their lifecycle

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system changes parameters by transforming data into encoded form with specific proportional distributions (e.g., 50% zeros and 50% ones) before storage. This parameter transformation ensures uniform wear across flash memory blocks and enables reliable detection of marginal blocks through their distinctive data distribution patterns

Inventive Principle:
Principle #35Parameter changes

2Volume of stationary object

If marginal blocks are not identified and tagged, then storage space is maximized, but data integrity and read reliability deteriorate due to bit errors in degraded blocks

Engineering Contradiction:
Improveusable storage spaceVSAvoiddata integrity
Core Design Contradiction:
Volume of stationary objectVSManufacturing precision

Solution Approach 1:

The system introduces an intermediary mechanism - the marginal tag with non-proportional data distribution - that acts as a mediator between the storage system and marginal blocks. This tag enables the system to identify and isolate marginal blocks without permanently losing the physical space, while ensuring data integrity by preventing writes to these blocks

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system applies local quality by treating marginal blocks differently from healthy blocks through the marginal tag mechanism. Instead of uniformly managing all blocks, the system selectively identifies and tags only those blocks exhibiting marginal characteristics, allowing differentiated management based on local block conditions

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS20130332796A1Storage control system with data management mechanism and method of operation thereof
Publication Date: 2013.12.12 SANDISK TECHNOLOGIES LLC
  • US20130332796A1 patent drawing
  • US20130332796A1 patent drawing
  • US20130332796A1 patent drawing

AI summary

A method of operation of a storage control system includes: generating encoded data having a proportional data distribution for writing to a memory device; identifying a marginal block when an erase block is read from the memory device; and generating a marginal tag for the marginal block, the marginal tag having a non-proportional data distribution different from the proportional data distribution.