Per-Row Count Based Refresh Target Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory devices face challenges in balancing targeted refresh operations with time and power consumption, particularly in identifying and refreshing memory cells affected by access patterns that increase information decay.
Innovation Solution
A memory device that includes a refresh control circuit which identifies rows with the highest access counts by storing count values associated with each word line and performing targeted refresh operations when the count value crosses a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If targeted refresh operations are performed on all rows to prevent information loss, then reliability is improved, but power consumption and time increase
Solution Approach 1:
The patent applies local quality by performing refresh operations selectively on specific rows that are identified as aggressor rows or victim rows, rather than uniformly refreshing all rows. The refresh control circuit identifies rows with high access counts and targets only those rows for refresh operations, making the refresh quality localized to where it is most needed.
Solution Approach 2:
The patent implements partial action by performing refresh operations on only a subset of rows (those with access counts above a threshold) rather than all rows. This partial refresh approach is sufficient to prevent information loss in critical rows while avoiding unnecessary power consumption on rows that do not require refreshing.
2Reliability
If targeted refresh operations are performed frequently to prevent information loss, then reliability is improved, but time consumption increases
Solution Approach 1:
The patent applies local quality by performing refresh operations selectively on specific rows that are identified as aggressor rows or victim rows, rather than uniformly refreshing all rows. The refresh control circuit identifies rows with high access counts and targets only those rows for refresh operations, making the refresh quality localized to where it is most needed.
Solution Approach 2:
The patent implements partial action by performing refresh operations on only a subset of rows (those with access counts above a threshold) rather than all rows. This partial refresh approach is sufficient to prevent information loss in critical rows while avoiding unnecessary time consumption on rows that do not require refreshing.
3Measurement precision
If access count tracking is implemented for all rows, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the necessary information (access counts) from row access operations and stores them in compact count storage locations. Rather than tracking detailed access patterns for all rows, the system extracts and stores only the essential count data needed to identify aggressor rows, simplifying the overall measurement system.
Solution Approach 2:
The patent changes the measurement parameter from detailed access pattern tracking to simple access count tracking. By monitoring only the number of accesses (a scalar parameter) rather than full access patterns, the system achieves sufficient measurement precision for identifying aggressor rows while significantly reducing device complexity.
Data Source
AI summary
Apparatuses and methods per-row count based refresh target identification. A memory device stores count values associated with the word lines. An aggressor detector circuit stores a maximum of the count values and a row address associated with the maximum count value. Responsive to a targeted refresh signal, the stored count value is compared to a threshold. If the count value has crossed the threshold, then a targeted refresh operation may be performed on one or more refresh addresses based on the stored address, and the count value may be reset.