Memory Read Threshold Adaptation via Single-Point Perturbation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory devices face challenges in accurately tracking and setting read thresholds due to variations in analog values over time and between memory cell groups, making it difficult to isolate the individual contribution of each read threshold to error metrics and decide on adjustments.
Innovation Solution
A method where only one read threshold is perturbed at a time during readout operations, allowing changes in error metrics to be attributed to its positioning, enabling the estimation of a preferred threshold value based on readout results, while keeping the other threshold fixed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple read thresholds are adjusted simultaneously based on error metrics, then the overall read accuracy may improve, but it becomes impossible to isolate the individual contribution of each threshold to the error metrics
Solution Approach 1:
The patent segments the threshold adjustment process by perturbing only one read threshold at a time while keeping other thresholds fixed. This segmentation allows the system to isolate and measure the individual contribution of each threshold to error metrics, resolving the contradiction between improving read accuracy and managing adjustment complexity.
Solution Approach 2:
Instead of adjusting all thresholds simultaneously (excessive action), the patent applies partial action by adjusting only one threshold at a time. This partial adjustment strategy enables clear attribution of error metric changes to specific threshold perturbations, simplifying the adjustment process while maintaining accuracy improvement capabilities.
2Measurement precision
If dedicated readout operations are performed to estimate preferred threshold values, then threshold accuracy improves, but additional readout operations increase time consumption
Solution Approach 1:
The patent merges the threshold estimation function with normal readout operations. By perturbing one threshold at a time during regular readout operations and using the resulting error metrics to estimate preferred threshold values, the system achieves accurate threshold estimation without requiring separate dedicated readout operations, thus eliminating additional time consumption.
Solution Approach 2:
Normal readout operations are given multiple functions: they serve both to retrieve data and to estimate preferred threshold values through controlled perturbation of a single threshold. This multi-functionality eliminates the need for dedicated threshold estimation operations, reducing time loss while maintaining estimation accuracy.
Data Source
AI summary
A method includes storing data in memory cells by programming the memory cells with respective values. The memory cells are read in multiple readout operations that each compares the programmed values to at least first and second read thresholds, while keeping the first read threshold fixed throughout the readout operations and perturbing only the second read threshold between the readout operations. A preferred value for the second read threshold is estimated based on the multiple readout operations.