Analog Memory Readout Using Compressed Soft Metrics for ECC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory devices face challenges in efficiently reading data from analog memory cells due to varying confidence levels of storage values, leading to potential read errors and high communication traffic when transferring confidence levels to Error Correction Code (ECC) decoders.
Innovation Solution
The method involves estimating confidence levels of storage values, compressing them, and transferring these compressed levels to a memory controller, where they are decompressed and used for ECC decoding, allowing for improved data readout performance with reduced communication traffic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If confidence levels are transferred to ECC decoder, then ECC decoding effectiveness is improved, but communication traffic increases
Solution Approach 1:
The patent extracts only the essential confidence level information needed for ECC decoding, rather than transferring all raw confidence data. This selective extraction reduces communication traffic while maintaining the effectiveness needed for soft decoding operations.
Solution Approach 2:
The patent transforms confidence levels into different parameter representations that are more compact. By changing the parameter format of confidence information, the system reduces the amount of data that needs to be communicated to the ECC decoder while preserving the essential decoding effectiveness.
2Measurement precision
If multiple read operations are performed to improve read accuracy, then data quality is improved, but read time increases
Solution Approach 1:
The patent performs preliminary read operations to gather confidence level information before the final ECC decoding step. By preparing confidence data in advance through initial read operations, the system improves subsequent decoding accuracy without requiring excessive additional read time during critical operations.
Solution Approach 2:
The patent maintains continuous useful action by integrating confidence level estimation into the read process itself, rather than performing separate additional read operations. The confidence information is gathered as part of the normal read operation, ensuring continuous data flow without interrupting the overall read timeline.
Data Source
AI summary
A method for data storage includes storing data in a group of analog memory cells by writing respective input storage values to the memory cells in the group. After storing the data, respective output storage values are read from the analog memory cells in the group. Respective confidence levels of the output storage values are estimated, and the confidence levels are compressed.The output storage values and the compressed confidence levels are transferred from the memory cells over an interface to a memory controller.