Cached Syndrome-Based Error Correction for NAND Flash
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
NAND flash memory devices face significant challenges in storage reliability due to noise sources like random-telegraph noise and inter-cell interference, which are exacerbated by aggressive technology scaling and multi-level per cell techniques, necessitating efficient error correction methods to maintain a non-recoverable bit error rate below 10−15.
Innovation Solution
Implementing a method to calculate and store error correction information in a cache based on syndromes, where the number of calculation cycles determines when to store the information, allowing for faster retrieval and reducing the need for time-consuming Berlekamp-Massey and Chien-search algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction information is calculated using Berlekamp-Massey and Chien-search algorithms, then accurate error correction is achieved, but the calculation time and processor intensity increase significantly
Solution Approach 1:
The patent pre-calculates and stores error correction information for various syndrome patterns in a cache memory before actual error correction is needed. When errors occur, the system first checks if the syndrome matches pre-calculated patterns in the cache, avoiding time-consuming real-time computation of Berlekamp-Massey and Chien-search algorithms while maintaining accurate error correction
Solution Approach 2:
The patent creates copies of frequently occurring error correction information and stores them in cache memory. Instead of recalculating the same error correction data repeatedly using complex algorithms, the system retrieves pre-computed copies from cache, significantly reducing calculation time while preserving correction accuracy
2Quantity of substance
If technology scaling and multi-level per cell techniques are used to reduce bit cost, then storage capacity increases, but storage reliability deteriorates due to increased noise sensitivity
Solution Approach 1:
The patent changes the parameter of error correction aggressiveness based on the syndrome pattern. For syndromes indicating minor errors (common in scaled devices), simpler and faster correction methods are applied. For syndromes indicating complex errors, the full Berlekamp-Massey and Chien-search algorithms are used. This adaptive approach maintains reliability while improving speed for the majority of cases
Data Source
AI summary
A computing device for correcting data errors may receive data stored by a memory device; calculate a syndrome associated with the data; initiate a calculation of error correction information for the data based on the syndrome; search for the error correction information in a cache based on the syndrome; discontinue the calculation of the error correction information when the error correction information is found in the cache before the error correction information is calculated; and correct an error associated with the data using the error correction information from the cache.