Memory Defect Detection via Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in accurately detecting defective memory devices due to correctable errors caused by sources other than memory device defects, such as random noise, which obscures identifying true defects in memory devices.
Innovation Solution
A system that filters rank error count values to suppress random noise-related errors, using a digital filter to distinguish between errors from memory device defects and other sources, allowing for timely maintenance before the device becomes non-operational.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If error counting is used to detect memory device defects, then defect detection capability is improved, but false positives from random noise increase
Solution Approach 1:
The patent extracts and removes random noise errors from the total error count by comparing errors detected during read operations against errors detected during self-refresh operations. By isolating the noise component and subtracting it from the total count, the system achieves more accurate defect detection without false positives from random noise
Solution Approach 2:
The system uses feedback from self-refresh operations to continuously characterize and update the random noise profile. This feedback mechanism allows the system to adapt to changing noise conditions and maintain accurate defect detection by comparing current error patterns against the established noise baseline
2Reliability
If monitoring continues until failure to enable maintenance, then system availability is improved, but detection timing precision deteriorates
Solution Approach 1:
The patent performs preliminary characterization of random noise behavior during self-refresh operations before actual memory operations. By establishing a baseline noise profile in advance, the system can immediately and accurately identify true defects without delay, enabling timely maintenance while maintaining high system availability
Data Source
AI summary
A technique includes receiving data indicative of a time varying count of errors, which are attributable to at least one memory device. The technique includes filtering the indicated count and detecting a defect in the memory device(s), where the detecting includes selectively generating an indicator to represent that the memory device(s) is defective based at least in part on a result of the filtering.