Flash Memory Copy Signature and Erased Flag Reuse
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing flash memory management systems face inefficiencies in identifying copied sectors without degrading write performance or consuming additional write operations, particularly in Multi-Level Cell (MLC) NAND flash devices with limited Partial Page Programming (PPP) capabilities.
Innovation Solution
Assigning a 'signature' to data chunks that remains consistent across copies, allowing identification of copied data without additional write operations, thereby enabling efficient reuse of storage space without the need for 'copy marks'.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If copy marks are used to identify copied sectors, then sector identification accuracy is improved, but write performance degrades due to additional write operations
Solution Approach 1:
The patent merges the copy identification function with the existing data structure by using the 'erased' flag in the flash memory page structure to indicate copied sectors. This eliminates the need for separate copy mark structures and avoids additional write operations, as the same flag is reused for both erase indication and copy completion indication.
Solution Approach 2:
The patent makes the 'erased' flag multi-functional: it serves dual purposes as both an erase indication and a copy completion marker. This universal usage allows the system to identify copied sectors without requiring separate dedicated markers, thereby avoiding additional write operations and maintaining write performance.
2Measurement precision
If copy marks are written for each sector, then copied sector identification is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple functions into the existing page structure: the 'erased' flag simultaneously indicates both erase operations and copy completion. This merging approach eliminates the need for separate copy mark structures, reducing device complexity while maintaining accurate copied sector identification.
Solution Approach 2:
The patent uses the existing data structure (page structure with erased flag) as a template for copy identification, rather than creating new complex mark structures. This copying of the existing structure's functionality allows copy detection without increasing device complexity.
3Measurement precision
If additional write operations are performed for copy marks, then copied sector tracking is improved, but Partial Page Programming availability is reduced
Solution Approach 1:
The patent merges copy tracking functionality into the existing page write operations by reusing the 'erased' flag. This eliminates the need for separate write operations to set copy marks, thereby preserving Partial Page Programming availability for actual data writes rather than metadata updates.
Solution Approach 2:
The system uses the existing page structure and flag mechanism to serve the copy tracking function without requiring additional write operations. The same infrastructure that handles erase operations also handles copy completion indication, making the system self-sufficient and preserving PPP resources for data operations.
Data Source
AI summary
Upon the arrival at a memory device of one or more data chunks associated with respective logical addresses, each data chunk is assigned a signature, stored in a first location, and copied to a second location. The copy is assigned a signature that matches the signature of its parent data chunk. Before erasing a memory block that includes one or more data chunks, it first is verified that those data chunks have been copied, i.e., that copies of all the data chunks in the block, with matching signatures, exist in the memory device.