Faulty Memory Block Grouping for Reliable Data Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage systems face increased fault rates due to reduced size and power consumption, particularly in on-chip caches, where memory cells fail more frequently at lower voltages and smaller sizes, and existing error correction methods rely on predictable error polarities that can change unpredictably.
Innovation Solution
The system groups data blocks with faulty bits together, ensuring at least one non-faulty bit per location, using a selector data store to indicate which bits are non-faulty, allowing for efficient storage and retrieval without discarding faulty blocks, and can adjust the number of blocks in a group based on error rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If memory cells are miniaturized and operated at lower voltages to reduce size and power consumption, then area and energy usage are improved, but fault rate increases
Solution Approach 1:
The patent divides a data block into multiple sub-blocks, where each sub-block is independently decodable. This segmentation allows the system to recover data even when some sub-blocks contain faults, thereby maintaining reliability while using miniaturized memory cells with higher fault rates
Solution Approach 2:
The patent changes the structural parameters of data storage by introducing sub-blocks with different fault tolerance characteristics. By adjusting the number and size of sub-blocks, the system can adapt to different fault rates caused by miniaturization while maintaining acceptable reliability
2Device complexity
If error correction methods rely on predictable error polarities and block pairing, then error correction is simpler, but the method becomes ineffective when error polarities are unpredictable
Solution Approach 1:
The patent performs preliminary encoding that creates redundant information in a way that does not depend on predicting error polarities. The sub-block structure is prepared in advance with inherent fault tolerance, allowing the system to handle unpredictable errors without complex runtime analysis
Solution Approach 2:
The patent extracts the dependency on error polarity prediction by designing an error correction mechanism that works regardless of error direction. The sub-block decoding approach separates the correction logic from error polarity assumptions, making the system effective for unpredictable errors
3Reliability
If faulty data blocks are discarded to ensure data integrity, then data accuracy is improved, but the fraction of usable memory blocks decreases
Solution Approach 1:
Instead of discarding entire faulty data blocks, the patent recovers usable data by dividing blocks into sub-blocks and selecting only the non-faulty sub-blocks for data reconstruction. This approach recovers a larger fraction of memory blocks while maintaining data integrity
Solution Approach 2:
By segmenting data blocks into smaller sub-blocks, the patent enables selective recovery of non-faulty portions. This segmentation allows the system to utilize faulty blocks partially, thereby increasing the fraction of usable memory while ensuring data integrity through selective sub-block decoding
Data Source
AI summary
Data storage control circuitry for controlling storage and retrieval of data in a data store in which data is stored in data blocks. A group data store stores data by grouping together blocks that have at least one faulty bit into groups of at least two blocks. For each group of blocks at least one of the blocks has a non-faulty bit for each of the bit locations in the blocks. A selector data store stores indicators for each group indicating which bits of the blocks within a group are the non-faulty bits. When storing data to a data block within a group, the data is stored in each of the blocks within the group. When retrieving data from a data block within a group, the data is read from respective bits of the blocks within the group as indicated by the indicators.