Memory Read Threshold Tracking Using Device-Originated Error Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory read voltage threshold calibration techniques are inaccurate and latency-prone, failing to effectively track voltage distribution shifts due to random workloads and operating conditions, leading to uncorrectable read data.
Innovation Solution
Utilizing memory device-originated metrics, such as failed byte count and failed bit count, to iteratively adjust read voltage thresholds, minimizing latency and ensuring accuracy by mapping these metrics to voltage threshold adjustments through lookup tables or mathematical transformations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing memory read voltage threshold calibration techniques are used, then the calibration process can be performed, but the accuracy is insufficient and latency is high
Solution Approach 1:
The memory device performs self-calibration by utilizing its own internal metrics (failed byte count, failed bit count) to adjust read voltage thresholds. This self-service approach eliminates the need for external calibration equipment and reduces calibration latency while improving accuracy through device-originated feedback.
Solution Approach 2:
The calibration process implements feedback mechanisms where the memory device monitors its own read operations, identifies failed bytes or bits, and uses this feedback information to iteratively adjust voltage thresholds. The feedback loop continuously refines calibration accuracy by comparing actual read outcomes with expected results and making real-time voltage adjustments.
2Reliability
If voltage threshold calibration is performed to improve read accuracy, then bit error rate decreases, but data transfer bandwidth is reduced
Solution Approach 1:
The system performs partial calibration by calibrating only specific voltage thresholds that are most critical for current operating conditions rather than all thresholds. This selective approach maintains sufficient read accuracy while minimizing the impact on data transfer bandwidth by reducing the time spent on calibration operations.
Solution Approach 2:
Voltage threshold calibration is performed periodically rather than continuously, allowing normal data transfer operations to proceed at full bandwidth between calibration cycles. The periodic calibration maintains reliability by updating voltage thresholds at appropriate intervals without constantly interrupting data transfer and reducing bandwidth.
Data Source
AI summary
Described are systems and methods for memory read threshold tracking based on memory device-originated metrics characterizing voltage distributions. An example memory device includes: a memory array having a plurality of memory cells and a controller coupled to the memory array. The controller is to receive one or more values of a metric characterizing threshold voltage distributions of a subset of memory cells. The controller is further to determine, via one or more calibration operations, one or more voltage threshold adjustment values based on the one or more values of the metric. The controller is further to apply the one or more voltage threshold adjustment values of reading the subset of memory cells.