Memory Cell Repair Using Failure-Count Backup Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Emerging memory devices such as FRAMs, MRAMs, and PCMs face challenges in maintaining reliability due to varying endurance and reliability of memory cells, leading to data storage failures, which conventional ECC methods struggle to effectively correct as data complexity increases.
Innovation Solution
Implementing an error table and repair table to monitor and replace memory cells with backup cells based on failure counts, ensuring dynamic error correction and efficient use of limited backup resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional ECC methods are used to correct data errors, then data integrity can be maintained to some extent, but the methods struggle to effectively correct errors as data complexity increases and memory cell reliability varies
Solution Approach 1:
The patent implements dynamic monitoring of memory cell failure counts and dynamically replaces cells based on observed failure patterns. The system continuously tracks which memory cells fail most frequently and replaces them with backup cells, adapting to changing reliability conditions rather than using static error correction codes. This dynamic approach allows the system to effectively handle varying data complexity and memory cell reliability.
Solution Approach 2:
The patent pre-allocates backup memory cells before failures occur and proactively replaces high-failure-count cells with these backup cells. By preparing replacement cells in advance and performing replacements before catastrophic failures occur, the system maintains data integrity more effectively than reactive ECC methods, especially as data complexity increases.
2Reliability
If backup memory cells are allocated to replace failed cells, then memory reliability improves, but the limited number of backup cells must be used efficiently
Solution Approach 1:
The patent changes the parameter of cell selection from random or uniform replacement to replacement based on failure count thresholds. By monitoring and comparing failure counts against dynamic thresholds, the system prioritizes replacement of the most problematic cells first, maximizing the impact of each backup cell allocation. This parameter-based selection ensures limited backup cells are used most efficiently to achieve the greatest reliability improvement.
Solution Approach 2:
The patent implements a process where backup cells are allocated to replace failed cells, and previously replaced cells can be recovered and reused as backup cells when their original backup is needed. This circular replacement strategy allows the limited number of backup cells to serve multiple cells over time, effectively managing the constraint on backup cell quantity while maintaining high memory reliability.
3Ease of manufacture
If memory cells with varying endurance are used, then manufacturing costs are reduced, but reliability maintenance becomes challenging
Solution Approach 1:
The patent applies local quality by treating each memory cell individually based on its observed failure characteristics rather than treating all cells uniformly. The system monitors failure counts for each cell and applies replacement decisions locally to specific high-failure cells. This allows the system to work effectively with memory cells of varying endurance, replacing only those that exhibit problematic failure patterns while leaving reliable cells in use, thereby maintaining overall reliability without requiring uniform high-endurance cells across the entire array.
Data Source
AI summary
A memory device includes: a memory cell array comprising a plurality of memory cells, the plurality of memory cells comprising a plurality of data memory cells including a first data memory cell and a plurality of backup memory cells including a first backup memory cell; a storage storing an error table configured to record errors in the plurality of data memory cells, the error table including a plurality of error table entries, each error table entry corresponding to one of the plurality of data memory cell and having an address and a failure count; and a controller configured to replace the first data memory cell with the first backup memory cell based on the error table.