Memory Read Threshold Calibration Using Decoded Data Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage systems using multi-level cells, variations in threshold voltage due to process variations, data retention issues, and program disturb conditions lead to inaccuracies in reading data, making default read thresholds unsuitable for devices that have experienced such conditions, and existing calibration methods like syndrome weight-based bit error rate estimation are noisy and complex.
Innovation Solution
A decision-based method for calibrating memory read thresholds using decoded data to directly minimize failed bit count, replacing syndrome weight with bit error rate for more accurate and efficient calibration, which can be implemented on a CMOS bonded array to reduce overhead and improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If syndrome weight-based bit error rate estimation is used for read threshold calibration, then calibration can be performed, but the method is noisy and complex
Solution Approach 1:
The patent changes the calibration parameter from syndrome weight to bit error rate (BER) directly. Instead of using syndrome weight as a proxy for BER, the invention directly minimizes BER by adjusting read thresholds based on decoded data feedback, thereby improving measurement precision while simplifying the calibration approach
Solution Approach 2:
The invention implements a feedback mechanism where decoded data from memory reads is used to calculate actual bit error rates, which then feed back into adjusting the read thresholds. This closed-loop feedback system enables more accurate calibration by directly measuring calibration quality through BER rather than using indirect syndrome weight metrics
2Ease of operation
If default read thresholds are used, then read operations are simple, but inaccuracies occur due to process variation, data retention issues, and program disturb conditions
Solution Approach 1:
The patent applies preliminary action by performing read threshold calibration before normal read operations. The calibration process pre-adjusts read thresholds based on actual memory characteristics and conditions, ensuring that subsequent read operations use optimized thresholds rather than default values, thereby improving reading accuracy without adding complexity to the read operation itself
3Reliability
If read threshold calibration is performed using traditional methods, then calibration is achieved, but latency is high and sequential read throughput is reduced
Solution Approach 1:
The invention enables continuous calibration by performing it in the background during idle periods or alongside normal operations rather than requiring a separate calibration phase. The calibration process uses decoded data from ongoing read operations to continuously refine thresholds, maintaining accuracy without interrupting the read throughput or adding significant latency
Data Source
AI summary
A read threshold voltage can vary over time due to process variation, data retention issues, and program disturb conditions. A storage system can calibrate the read threshold voltage using data from a decoded codeword read from a wordline in the memory. For example, the storage system can use the data instead of syndrome weight in a bit error rate estimate scan (BES). As another example, the storage system can use the data to generate a bit error rate distribution, which can be used instead of a cell voltage distribution histogram. Using these techniques can help reduce latency and power consumption, increase throughput, and improve quality of service.