Memory Testing Method Using Multi-Grade Classification for Transient Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contemporary computing systems face failures due to defective memory cells, which can lead to system crashes and loss of critical information, as existing memory protection techniques only distinguish between 'good' and 'bad' memory blocks without accounting for potentially weaker cells that may not cause immediate failure.
Innovation Solution
A two-part testing method is employed, consisting of a primary test to identify potentially defective memory cells and a focus test to verify their defects, allowing for multiple quality grades of memory blocks and enabling the use of memory blocks with transient errors, thereby improving memory utilization and reducing the risk of system crashes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If memory blocks are classified only as good or bad, then memory protection is simple and fast, but memory utilization is reduced because potentially usable blocks with transient errors are discarded
Solution Approach 1:
The patent applies local quality by introducing multiple quality grades (first grade, second grade, third grade) for different memory blocks based on their error characteristics. Instead of uniform good/bad classification, memory blocks are differentiated locally according to their specific error patterns, allowing potentially usable blocks to be retained and assigned appropriate usage restrictions based on their grade.
Solution Approach 2:
The patent changes the classification parameter from a binary good/bad state to a multi-grade quality state. By introducing quality grades that reflect different error characteristics (permanent errors vs. transient errors), the system transforms the memory classification parameter to enable more nuanced decision-making about memory block usability, thereby increasing overall memory utilization while maintaining reliability.
2Stability of the object's composition
If all memory blocks with errors are discarded, then system stability is maintained, but memory resource waste increases
Solution Approach 1:
The patent applies discarding and recovering by selectively discarding only those memory blocks with permanent errors (third grade) while recovering and retaining memory blocks with transient errors (first and second grades) for continued use. This selective approach allows the system to discard problematic blocks while recovering usable capacity from blocks that can still function reliably under certain conditions.
Solution Approach 2:
The patent introduces dynamic classification where memory block quality grades can change over time based on error patterns. Memory blocks are not statically labeled as good or bad but are dynamically assessed and reclassified based on their error behavior, allowing the system to adapt to changing memory characteristics and optimize resource utilization while maintaining stability.
3Measurement precision
If comprehensive verification testing is performed on all memory blocks, then defect detection accuracy is high, but testing time and system performance degradation are significant
Solution Approach 1:
The patent applies partial action by performing comprehensive verification testing only on memory blocks that fail the initial screen test (those requiring quality grade determination), rather than testing all memory blocks exhaustively. This partial application of rigorous testing reduces overall testing time while maintaining high detection accuracy for problematic blocks, balancing thoroughness with efficiency.
Solution Approach 2:
The patent segments the testing process into two distinct stages: an initial screen test that quickly identifies potentially problematic blocks, and a subsequent verification test that performs comprehensive checking only on those blocks. This segmentation of the testing workflow allows the system to process large volumes of memory blocks efficiently while maintaining high accuracy in defect detection for the blocks that require it.
Data Source
AI summary
Embodiments of systems and methods for testing memory are disclosed, where memory errors are detected, and, in at least one embodiment, memory units containing errors are prevented from being accessed by applications on a computing system.