Non-volatile Memory Compression Engine Paired Pages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing non-volatile semiconductor memory systems face inefficiencies in data storage and retrieval due to limitations in single-level cell (SLC) and multi-level cell (M) technologies, particularly in terms of performance, endurance, and capacity, which are not optimally addressed by current memory management protocols.
Innovation Solution
The implementation of a compression engine within the memory device that dynamically switches between S-mode and M-mode storage based on data size, using paired pages to optimize storage efficiency by compressing write data and decompressing read data, thereby enhancing performance and endurance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If SLC mode is used for storage, then performance and endurance are improved, but capacity is reduced
Solution Approach 1:
The memory device divides its storage pages into paired groups, where each pair consists of two pages. This segmentation allows the system to selectively use SLC mode for one page and MLC mode for the other page in each pair, enabling a balance between performance (SLC characteristics) and capacity (MLC characteristics) at the page level rather than requiring the entire device to operate in a single mode.
2Quantity of substance
If MLC mode is used for storage, then capacity is increased, but performance and endurance are reduced
Solution Approach 1:
The system applies different storage modes to different local regions (pages) based on data characteristics. Specifically, highly compressible data is stored in SLC mode within paired pages to achieve high performance, while less compressible data is stored in MLC mode to maximize capacity utilization. This local quality approach allows each page to have the optimal storage mode for its specific data type.
3Quantity of substance
If data is compressed before storage, then storage efficiency is improved, but access latency increases
Solution Approach 1:
The compression engine performs data compression in advance during write operations, so that compressed data is already prepared and stored in SLC mode before read operations occur. This preliminary compression action eliminates the need for real-time compression during read operations, thereby minimizing access latency while maintaining high storage efficiency.
Data Source
AI summary
A non-volatile semiconductor memory is disclosed comprising a plurality of paired pages, wherein each pair comprises a first page and a second page. A write command is received from a host comprising write data and a write address. The write address is mapped to a physical address of a selected one of the paired memory pages. The write data is compressed to generate compressed data, and when the compressed data fits in one of the pages of the selected pair, the compressed data is stored in the first page and in the second page of the selected pair in an S-mode.


