Non-volatile Memory Wear Prediction Using Fermi-Dirac Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current non-volatile memory systems face challenges in efficiently managing and predicting the failure of memory blocks due to high error rates, which can lead to data corruption and reduced system reliability, especially when dealing with low-probability events that require significant time and resources for accurate prediction.
Innovation Solution
The implementation of a statistical analysis method using a Fermi-Dirac distribution to extrapolate from a small sample size, allowing for rapid and cost-effective prediction of low-probability events such as target Failed Bit Counts (FBC) in non-volatile memory systems, enabling proactive management of wear leveling, garbage collection, and voltage adjustments within the memory system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional statistical methods are used to predict memory block failures, then measurement precision can be achieved with large sample sizes, but time consumption and computational resources increase significantly
Solution Approach 1:
The patent changes the mathematical parameters and approach by using Fermi-Dirac distribution instead of traditional statistical methods. This allows the system to achieve accurate predictions of low-probability events (memory block failures) by transforming the problem into a different mathematical framework that requires significantly fewer samples and computational resources while maintaining prediction accuracy.
2Measurement precision
If traditional statistical methods are used to predict memory block failures, then measurement precision can be achieved, but computational resources and cost increase significantly
Solution Approach 1:
The patent changes the mathematical parameters and approach by using Fermi-Dirac distribution instead of traditional statistical methods. This allows the system to achieve accurate predictions of low-probability events (memory block failures) by transforming the problem into a different mathematical framework that requires significantly fewer samples and computational resources while maintaining prediction accuracy.
3Reliability
If wear leveling and garbage collection are performed frequently to manage memory block failures, then system reliability improves, but productivity and operational efficiency decrease
Solution Approach 1:
The patent applies preliminary action by predicting which memory blocks are likely to fail before they actually do, using the Fermi-Dirac distribution to estimate failure probabilities based on current wear patterns. This allows the system to perform wear leveling and garbage collection proactively on blocks that are most at risk, rather than reactively after failures occur, thereby maintaining reliability while reducing unnecessary operations and improving overall productivity.
Data Source
AI summary
A non-volatile storage apparatus includes a set of non-volatile memory cells and one or more control circuits in communication with the set of non-volatile memory cells. The one or more control circuits are configured to collect failure bit counts (FBCs) for data read from the set of non-volatile memory cells in a first time period and manage the set of non-volatile memory cells according to a probability of occurrence of a target FBC in a second time period that is subsequent to the first time period. The probability of occurrence of the target FBC during the second time period is calculated from a model of FBC distribution change of the set of non-volatile memory cells.