Selective Data Pattern Write Scrub for Memory Subsystems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory sub-systems face issues with media degradation and inaccurate data readings due to voltage threshold drift in non-volatile memory devices, which are exacerbated by frequent scrubbing operations that rewrite all memory cells, leading to increased bandwidth consumption and reduced lifespan.
Innovation Solution
A selective data pattern write scrub method is implemented, where memory cells storing logic '0' are scrubbed less frequently, with a counter tracking the number of scrub cycles to determine when both logic states need to be rewritten, thereby reducing media degradation and bandwidth utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If frequent scrubbing operations are performed to correct voltage threshold drift, then data reading accuracy is improved, but media degradation increases and device lifespan decreases
Solution Approach 1:
The patent applies local quality by differentiating scrubbing frequency based on data pattern characteristics. Instead of uniformly scrubbing all memory cells, the system identifies specific data patterns (e.g., alternating 0s and 1s) that are more susceptible to voltage threshold drift and applies scrubbing selectively to those patterns, while reducing or skipping scrubbing for other patterns. This localized approach maintains reading accuracy for vulnerable data while reducing overall media degradation.
Solution Approach 2:
The patent implements partial action by performing scrubbing operations only on a subset of memory cells containing problematic data patterns rather than all memory cells. The system detects specific patterns (such as multi-bit transitions) and applies scrubbing only to those instances, leaving other memory cells untouched during that scrub cycle. This partial approach sufficient addresses voltage threshold drift issues without subjecting all media to unnecessary write cycles.
2Reliability
If frequent scrubbing operations are performed to maintain data integrity, then data reliability is improved, but bandwidth consumption increases
Solution Approach 1:
The patent applies local quality by directing scrubbing operations only to memory cells containing specific problematic data patterns rather than uniformly processing all memory cells. The system identifies patterns with higher susceptibility to voltage threshold drift (such as alternating bit patterns) and concentrates scrubbing resources on those specific locations, thereby maintaining data integrity for vulnerable data while minimizing bandwidth consumption on the overall memory subsystem.
Solution Approach 2:
The patent implements partial action by performing scrubbing operations on only a portion of memory cells during each scrub cycle. Instead of reading and rewriting all memory cells, the system selectively targets cells containing identified problematic patterns. This partial approach maintains sufficient data integrity for the affected patterns while significantly reducing the bandwidth consumption associated with comprehensive scrubbing of entire memory blocks.
3Measurement precision
If all memory cells are rewritten during scrubbing to correct voltage threshold drift, then data accuracy is improved, but device lifespan is reduced
Solution Approach 1:
The patent applies local quality by differentiating between memory cells based on their stored data patterns and applying scrubbing operations selectively. Memory cells containing problematic patterns (such as alternating 0s and 1s that are more prone to voltage threshold drift) are rewritten during scrubbing, while cells with other patterns are left untouched. This localized approach maintains data accuracy for vulnerable patterns while reducing the cumulative write cycles on the overall device, thereby extending lifespan.
Solution Approach 2:
The patent implements partial action by performing rewrite operations on only a subset of memory cells during each scrub cycle rather than all cells. The system identifies specific patterns that benefit from scrubbing and limits rewrite operations to those instances. This partial rewriting approach maintains sufficient data accuracy for the affected patterns while significantly reducing the total number of write cycles experienced by the device, thereby preserving media longevity.
Data Source
AI summary
A system includes a memory device having a plurality of memory cells and a processing device operatively coupled to the memory device. The processing device is to determine to perform a rewrite on at least a portion of the plurality of memory cells. The processing device can determine that a number of rewrite operations at first subset of memory cells storing a first logic state fail to satisfy a threshold criterion. The processing device can also cause a rewrite of data stored at a second subset of memory cells storing a second logic state in response to determining the number of rewrite operations performed at the first subset of memory cells fail to satisfy the threshold criterion.