Stream Data Update Layout for Low-Write-Amplification Flash Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory storage devices face inefficiencies in updating stream data due to high write amplification and stranded capacity when using non-volatile flash memory, particularly in protocols like NVMe, which lead to slow operation speeds, reduced longevity, and unnecessary power consumption.
Innovation Solution
The method involves writing stream data updates to a new superblock separate from the existing data, with additional data being smaller than a substream, and concurrently writing non-stream data to the same superblock, followed by reconciliation based on stream IDs or logical block addresses without modifying the existing data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If stream data updates are written to the same superblock as existing data, then data consistency is maintained, but write amplification increases and service lifetime decreases
Solution Approach 1:
The patent segments the superblock into distinct regions: a first region for storing original stream data and a second region for storing update data. This segmentation allows updates to be written without modifying the original data structure, reducing write amplification and extending service lifetime while maintaining data consistency through logical reconciliation.
2Productivity
If stream data updates are written to a new superblock, then write amplification is reduced, but storage capacity utilization decreases due to stranded capacity
Solution Approach 1:
The patent merges multiple update operations into a single superblock by writing multiple stream data updates and non-stream data concurrently to the second region. This consolidation reduces the number of superblocks needed, minimizing stranded capacity while maintaining low write amplification.
3Reliability
If stream data updates are processed sequentially, then data integrity is ensured, but operation speed decreases
Solution Approach 1:
The patent performs preliminary actions by writing update data to the second region concurrently with writing new stream data to the first region, before reconciliation is needed. This parallel processing approach maintains data integrity through subsequent reconciliation while significantly improving operation speed.
4Speed
If additional power is provided to accelerate write operations, then operation speed increases, but power consumption increases
Solution Approach 1:
The patent maintains continuous useful action by overlapping write operations - while stream data is being written to the first region, update data is concurrently written to the second region. This continuous parallel processing improves operation speed without requiring additional power bursts, thus reducing overall power consumption.
Data Source
AI summary
In accordance with some embodiments of the present disclosure, a method is performed on a memory storage device for updating data stored in the memory storage device. The method includes writing a first stream data to a first portion of memory, wherein the first stream data includes a plurality of substreams, and writing first additional data to a second portion of memory, wherein the first additional data includes data indicative of an update of at least one substream of the first stream data and is no larger than a substream of the plurality of substreams of the first stream data. Concurrent with writing the first additional data, the method also includes writing second additional data to the second portion of memory, wherein the second additional data includes at least one of data indicative of an update of at least one substream of a second stream data, or non-stream data.