Solid-State Storage Data Compression and Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices have limited erase/program cycles and high cost per bit of storage, necessitating improved management to maximize memory usage and extend device lifespan.
Innovation Solution
Implementing data collection and compression methods within a solid-state storage device, where partially programmed memory pages are identified, valid data is read, compressed, and stored in new memory blocks, allowing for efficient reuse of memory areas and reducing write amplification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If flash memory devices are used for data storage, then high memory density and reliability are achieved, but the device lifespan is limited due to restricted erase/program cycles
Solution Approach 1:
The patent changes the data organization parameters by implementing data collection that groups valid data from multiple partially programmed pages into a single full page, and applies compression to reduce data volume. This transforms how data is stored and managed, extending device lifespan by reducing erase/program cycle frequency while maintaining reliability through proper data validation and error correction codes
Solution Approach 2:
The patent recovers and reuses valid data from partially programmed pages that would otherwise be wasted. By identifying and collecting valid data from these partial pages, compressing it, and storing it efficiently in full pages, the system maximizes memory utilization and reduces the frequency of erase/program operations, thereby extending device lifespan
2Reliability
If flash memory devices are used for data storage, then non-volatile memory capability is achieved, but the cost per bit is high compared to other memory technologies
Solution Approach 1:
The patent changes the storage efficiency parameters through compression algorithms that reduce the number of bits required to store the same information. By compressing valid data before storage and implementing efficient data collection from partial pages, the system increases effective storage capacity, thereby reducing cost per bit while maintaining non-volatile memory capabilities
Solution Approach 2:
The patent makes the flash memory device perform multiple functions: data storage, data compression, data collection from partial pages, and error correction. This multi-functionality increases the effective utility of the memory device, providing better value and reducing the effective cost per bit by maximizing the use of each stored bit
3Adaptability or versatility
If data is stored in partially programmed memory pages, then memory writing flexibility is improved, but memory capacity utilization decreases and write amplification increases
Solution Approach 1:
The patent recovers valid data from partially programmed pages that would otherwise represent wasted storage capacity. By collecting this valid data, compressing it, and storing it in full pages, the system eliminates the capacity waste associated with partial pages while maintaining writing flexibility through the ability to handle incremental and variable data writes
Solution Approach 2:
The patent merges valid data from multiple partially programmed pages into a single consolidated full page after compression. This combining process maximizes memory capacity utilization by eliminating fragmented partial pages, while the compression step ensures that the merged data fits efficiently into the consolidated page structure
Data Source
AI summary
Memory systems may include a storage device comprising a first memory; a processor; and a second memory. The memory systems may be configured to read first data from a first storage location of the first memory, store the first data to a first storage location of the second memory and input the first data to a compression block of the storage device, compress the first data and store the compressed first data to a second storage location of the second memory, evaluate the compressed first data, store the first data from the first storage location of the second memory to a different storage location of the first memory when the compressed first data is not deemed compressible, and combine the compressed first data with additional compressed data and store the combined compressed data to a different storage location of the first memory when the compressed first data is deemed compressible.