NAND Flash BER Estimation for Adaptive SSD Read Flows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Solid-state drives (SSDs) face increased latency and reduced quality of service due to high bit error rates and inter-cell interference, which are exacerbated by high memory cell density and wear and tear, leading to inefficiencies in read operations and error correction processes.
Innovation Solution
The method involves estimating and classifying the bit error rate (BER) of memory cells using threshold voltage distributions, allowing for adaptive read flows that skip or modify decoding stages based on BER thresholds, thereby optimizing read operations and reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction codes (ECC) are used to correct read errors in NAND flash memories, then reliability is improved, but read latency increases
Solution Approach 1:
The patent applies partial action by selectively executing ECC decoding only when necessary. The controller estimates BER based on threshold voltage distribution and only performs hard decision decoding when the estimated BER exceeds a threshold, avoiding unnecessary decoding operations and reducing read latency while maintaining reliability.
Solution Approach 2:
The patent changes the parameter of decoding execution from always-executing to conditionally-executing based on BER estimation. By dynamically adjusting whether to perform ECC decoding based on the estimated bit error rate, the system optimizes the balance between reliability and read latency.
2Reliability
If multiple reads at varying threshold voltages are performed to improve read quality, then error correction capability is improved, but read time increases severely
Solution Approach 1:
The patent applies partial action by performing multiple threshold voltage reads only when the estimated BER indicates high error probability. When BER is low, the system performs a single read operation, significantly reducing read time while maintaining read quality when needed.
Solution Approach 2:
The patent performs preliminary BER estimation based on threshold voltage distribution before executing multiple reads. This preliminary action allows the system to decide in advance whether multiple reads are necessary, avoiding unnecessary read operations and reducing overall read time.
3Reliability
If hard decision decoding stages are executed to correct errors, then reliability is improved, but read flow latency is increased
Solution Approach 1:
The patent changes the execution parameter of hard decision decoding from mandatory to conditional based on BER classification. By dynamically adjusting decoding execution based on estimated error rates, the system maintains reliability when errors are likely while minimizing latency when errors are unlikely.
4Quantity of substance
If high memory cell density is used to increase storage capacity, then storage capacity is improved, but inter-cell interference increases leading to higher error rates
Solution Approach 1:
The patent implements feedback by using the estimated BER (derived from threshold voltage distribution affected by inter-cell interference) to dynamically adjust the read flow and decoding strategy. This feedback mechanism allows the system to compensate for errors caused by high-density inter-cell interference, maintaining reliability despite increased storage capacity.
Data Source
AI summary
A method for reading data from an SSD, comprising: retrieving data from a target row of memory cells using initial threshold voltages; decoding the data using a first hard decision decoding stage; estimating a bit error rate (BER) of a target row of memory cells based on a distribution of threshold voltages of cells in a memory block containing the target row when the first hard decision decoding stage fails; classifying the BER of the target row based on a first BER threshold (BER-TH1); and executing a first read flow comprising at least one hard decision decoding stage if the BER is less than the BER-TH1, and executing a second read flow similar to the first read flow if the BER is greater than or equal to the BER-TH1, the second read flow skipping a hard decision decoding stage of the first read flow.