SSD Data Block Remapping for Stronger First-Tier ECC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage devices face limitations in error correcting code capability, particularly when errors exceed the correction capability of the error correcting code, leading to unsuccessful corrections and increased latency due to the need for secondary error correcting code operations.
Innovation Solution
A remix operation is performed to change the logical to physical association of a data block, redistributing errors to allow the first tier of error correcting code operations to correct previously uncorrectable errors, thereby reducing the frequency of secondary error correcting code operations and latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional error correcting code operations are used, then errors within correction capability are corrected, but errors exceeding correction capability remain uncorrected and require secondary operations
Solution Approach 1:
The patent applies preliminary action by performing a remix operation before error correcting code operations. This remix operation pre-distributes data across multiple physical locations in a manner that anticipates potential error patterns, allowing the first tier ECC operations to correct errors that would otherwise require secondary operations. The remix operation prepares the data layout in advance to facilitate more effective error correction.
Solution Approach 2:
The patent employs dimensionality change by transforming the data layout from a conventional single-dimension arrangement to a multi-dimensional distributed arrangement across different physical locations. This remix operation creates a new spatial dimension in data organization, allowing errors to be distributed across multiple codewords rather than concentrated in a single codeword, thereby enabling the first tier ECC to handle more error scenarios.
2Reliability
If secondary error correcting code operations are performed, then previously uncorrectable errors are corrected, but computational resources and latency increase
Solution Approach 1:
The remix operation is performed in advance to reorganize data into a layout that maximizes the effectiveness of first-tier ECC operations. This preliminary reorganization ensures that when errors occur, they are more likely to be distributed in a manner that the first tier ECC can handle, thereby avoiding the need for computationally intensive secondary ECC operations.
Solution Approach 2:
The remix operation creates a redistributed copy of the data across different physical locations and codewords. Instead of attempting to correct concentrated errors using complex secondary operations, the system creates a copied and redistributed version of the data where errors are spread out, allowing simpler first-tier ECC operations to succeed.
3Ease of manufacture
If data is stored with fixed logical to physical associations, then storage structure is simple, but error distribution cannot be optimized for correction
Solution Approach 1:
The patent introduces dynamics by implementing a remix operation that can change the logical to physical association of data based on error patterns and correction needs. Rather than maintaining a fixed static mapping, the system dynamically reorganizes data layouts to optimize error distribution, allowing the storage structure to adapt to different error scenarios while maintaining manageable complexity through automated remapping procedures.
Data Source
AI summary
An instance of an event associated with error correcting code operations performed on a data block of the non-volatile memory is identified. An entry for a record is generated. The entry is indicative of the instance of the event. Whether a frequency of the event satisfies a threshold condition based on the record is determined. Responsive to determining that the frequency of the event satisfies the threshold condition, a remix operation on the data block is performed to change a logical to physical association of the data block from a first logical association to a second logical association.