Memory Error Correction Using Segmented ECC for Recurring Faults
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory devices, such as DRAM, experience recurring errors due to charge loss, which consume finite error correction code resources and can lead to memory failure when conventional error correction schemes can only correct single errors, leaving recurring errors undetected and uncorrected.
Innovation Solution
The method involves determining and correcting recurring errors in memory cells using a second subset of memory cells that stores error correction code and recurring error data, allowing for the reduction of error correction scheme resources and enabling the correction of additional errors without relying on conventional ECC schemes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional error correction code schemes are used to correct recurring errors, then error correction capability is provided, but finite ECC resources are consumed and additional errors cannot be corrected
Solution Approach 1:
The patent segments error correction into two distinct parts: a first error correction code for correcting single errors, and a second error correction code specifically for correcting recurring errors. This segmentation allows each code to be optimized for its specific error type, improving overall correction efficiency while managing ECC resources effectively.
Solution Approach 2:
The patent introduces an intermediary mechanism - a recurring error data structure that tracks and identifies recurring error patterns. This intermediary allows the system to detect and correct recurring errors before they consume the primary ECC resources, enabling the first error correction code to focus on non-recurring errors.
2Reliability
If larger error correction code resources are allocated to handle recurring errors, then more errors can be corrected, but device size and complexity increase
Solution Approach 1:
By dividing error correction into specialized segments (first ECC for single errors, second ECC for recurring errors), the patent avoids the need for a single complex comprehensive error correction system. Each segment can be simpler and more efficient at its specific task.
Solution Approach 2:
The system performs preliminary detection and correction of recurring errors using the second error correction code and recurring error data structure. By addressing recurring errors in advance, the system prevents them from overwhelming the primary ECC resources, simplifying the overall error correction burden.
3Device complexity
If conventional ECC schemes are used without recurring error detection, then implementation is simpler, but recurring errors remain undetected and uncorrected leading to memory failure
Solution Approach 1:
The patent implements preliminary detection of recurring errors through the second error correction code and recurring error data structure before these errors can cause memory failure. This preliminary action allows the system to identify and correct recurring error patterns early, preventing them from escalating into critical failures.
Solution Approach 2:
The system employs feedback mechanisms through the recurring error data structure that tracks error patterns and provides information about recurring errors. This feedback loop enables continuous monitoring and correction of recurring errors, improving memory reliability over time.
Data Source
AI summary
The present disclosure includes apparatuses and methods for correcting recurring errors in memory. A number of embodiments include determining whether a first subset of a group of memory cells has a recurring error associated therewith using a second subset of the group of memory cells, and responsive to a determination that the first subset of the group of memory cells has a recurring error associated therewith, correcting the recurring error using the second subset of the group of memory cells.