Flash Controller Erase Reattempt for Misidentified Bad Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash storage devices experience back-to-back erase failures due to block-to-block disturb effects, leading to misidentification of good blocks as 'grown bad blocks' (GBBs), which can result in the device entering a read-only mode, reducing performance and increasing testing costs.
Innovation Solution
A controller is configured to program an open block and attempt to erase multiple closed blocks, re-attempting erasures when consecutive failures occur and unmarking successfully erased blocks as good, while identifying and marking the aggressor block as a GBB to prevent further access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the controller marks closed blocks as bad blocks after erase failures, then the reliability of identifying truly defective blocks is improved, but good blocks are misidentified as bad blocks leading to reduced storage capacity and potential read-only mode
Solution Approach 1:
The controller performs a reattempt to erase closed blocks that were previously marked as bad blocks, before finally giving up on them. This preliminary reattempt action allows the system to recover from temporary erase failures and prevent misidentification of good blocks as permanent bad blocks, thereby maintaining storage capacity while still identifying truly defective blocks.
2Reliability
If the controller enters read-only mode to prevent further data writes, then data integrity is protected, but device performance is reduced and testing costs increase
Solution Approach 1:
Before entering read-only mode, the controller performs a preliminary check by reattempting to erase blocks that caused consecutive erase failures. This preliminary action distinguishes between temporary erase failures (which can be recovered) and permanent block defects (which require read-only mode), thereby preventing unnecessary performance degradation while still protecting data integrity when truly needed.
3Reliability
If the controller performs multiple erase attempts on closed blocks, then the likelihood of recovering misidentified good blocks is improved, but the time and complexity of erase operations increases
Solution Approach 1:
The controller performs a limited number of reattempt erases (a partial action) on previously failed blocks before giving up and marking them as permanent bad blocks. This partial reattempt approach recovers some misidentified good blocks without performing excessive erase operations that would waste time, achieving a balance between reliability and time efficiency.
Data Source
AI summary
Aspects of a storage device including a controller are provided which recovers misidentified bad blocks that fail to erase due to charge leakage from a previously programmed open block. The controller programs an open block, and attempts to erase a plurality of closed blocks following the programming of the open block. When the closed blocks fail to erase, the controller marks the closed blocks as bad blocks. The controller then determines whether a number of consecutive erase failures after programming the open block meets a threshold, in response to which the controller resets a die including the closed blocks and reattempts to erase the closed blocks. The controller then unmarks as bad blocks the closed blocks which successfully erased in response to the re-attempt.