Non-Volatile Memory Controller Log-Structured Write Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices face challenges in improving random write performance due to the significant difference between data written size and data read size, leading to increased processing time and deteriorated performance.
Innovation Solution
The implementation of a log-structured method, where write object data is sequentially written in ascending order of pages, regardless of the logical address, and compaction is performed to manage invalidated pages and free blocks, thereby optimizing data storage and retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If block managing method is used with integral multiple of block size as management size, then data storage management is simplified, but random write performance deteriorates due to significant difference between data written size and data read size
Solution Approach 1:
The patent segments the storage management into two independent units: a log structure for write operations that sequences data by physical address, and a mapping table for read operations that translates logical addresses to physical addresses. This segmentation allows write and read operations to be optimized independently, resolving the contradiction between simplified management and write performance.
Solution Approach 2:
The patent introduces a mapping table as an intermediary layer between the host device and the log structure. The mapping table stores correspondence between logical addresses and physical addresses, enabling efficient random read access without requiring data to be physically reorganized, thus maintaining both management simplicity and write performance.
2Reliability
If data is read and merged in temporary storage area before writing to new blocks, then data integrity is maintained, but process time increases significantly
Solution Approach 1:
The patent performs preliminary actions by pre-allocating new blocks and preparing the log structure before actual write operations. The log structure sequences write object data in advance by physical address, and the mapping table is updated beforehand to reflect new address mappings, reducing the time required during actual write operations while maintaining data integrity through structured validation.
3Productivity
If semiconductor storage element is used with page and block units, then storage efficiency is improved, but random write capability is disabled
Solution Approach 1:
The patent implements a dynamic address translation mechanism where the mapping table can be updated in real-time to reflect new logical-to-physical address mappings. This dynamic adaptation allows the system to maintain the efficient page and block structure of semiconductor storage while enabling random write operations through flexible address remapping without requiring physical data movement.
Data Source
AI summary
According to one embodiment, a write instructing unit instructs a data access unit to write, in a storage area of a data storage unit indicated by a first physical address, write object data, instructs a management information access unit to update address conversion information, and instructs a first access unit to update the first physical address. A compaction unit extracts a physical address of compaction object data, instructs the data access unit to read the compaction object data stored in a storage area of the data storage unit indicated by the physical address, instructs the data access unit to write the compaction object data in a storage area of the data storage unit indicated by a second physical address, instructs the management information access unit to update the address conversion information, and instructs a second access unit to update the second physical address.