Backup Buffer Compression for Non-Volatile Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The limited size of the Backup buffer in non-volatile memory devices restricts the number of secure and non-atomic update operations, leading to potential overflow and reduced lifetime due to frequent erase/write operations, as existing methods do not efficiently manage memory regions and erase operations.
Innovation Solution
A compression method that reuses Backup buffer entries by marking and erasing address entries cyclically, using a driver buffer to manage address pages and optimize storage within the Backup buffer, ensuring circular use and minimizing unnecessary operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If persistent data is copied to a Backup buffer during multiple writing operations, then data consistency is guaranteed after unexpected events, but the Backup buffer size is limited and frequent erase/write operations reduce its lifetime
Solution Approach 1:
The patent performs preliminary actions by marking address entries as erasable immediately after update operations complete, rather than waiting for buffer overflow. This allows the system to prepare erasure operations in advance, extending the usable lifetime of the Backup buffer while maintaining data consistency guarantees.
Solution Approach 2:
The patent implements a mechanism where address entries in the Backup buffer are marked as erasable after their associated update operations complete. This allows the system to recover buffer space by erasing only the necessary portions, rather than performing full buffer erasures, thereby extending the Backup buffer's operational lifetime while maintaining reliability.
2Quantity of substance
If the Backup buffer is frequently erased and rewritten to manage limited size, then buffer overflow is prevented, but the number of erase/write operations increases reducing overall system lifetime
Solution Approach 1:
The patent changes the state parameter of address entries from 'valid' to 'erasable' after update operations complete. This parameter change allows the system to manage buffer capacity dynamically without frequent full erasures, reducing wear on the Backup buffer and extending system lifetime while preventing overflow.
Solution Approach 2:
The system performs preliminary marking of address entries as erasable before actual erasure operations are needed. This allows buffer management to occur in advance, reducing the frequency and intensity of erase/write cycles while maintaining proper buffer capacity control.
3Reliability
If all memory portions involved in update operations are preserved in the Backup buffer, then data consistency is maintained, but the Backup buffer fills up faster requiring more frequent erasures
Solution Approach 1:
The patent allows the system to discard (mark as erasable) address entries in the Backup buffer after their associated update operations have successfully completed. This recovery of buffer space enables continuous update operations without frequent full buffer erasures, improving throughput while maintaining data consistency through the marking mechanism.
Solution Approach 2:
The system changes the status parameter of Backup buffer entries from 'protected' to 'erasable' based on transaction completion status. This allows selective buffer management where only completed transactions can have their entries erased, maintaining consistency for active transactions while freeing space for new operations.
Data Source
Figure 1~2a
Figure 2b~3
Figure 4a~4i
AI summary
Compression method for a Backup Data buffer (2), comprising a plurality of Backup entries (2a, 2b, 2c, 2d) for storing persistent data (data-R1, data-R2, data-R3, data-R4) of a non volatile memory (1) device during at least one update operation (op1, op2), an address of the persistent data (data-R1, data-R2, data-R3, data-R4) in the non volatile memory (1) device being stored in a driver buffer (drv) comprising a plurality of address pages (pag1, pagm), each including a plurality of address entries (A1, An), the compression method comprising the functions for: marking as erasable an address entry (A1), included in a first address page (pag1) of the driver buffer (drv), when the at least one update operation (op1, op2) on the persistent data (data-R1, data-R2, data-R3, data-R4) is completed; copying, from the first address page (pag1) to a second address page (pag2) of the driver buffer (drv) a plurality of address entries (A2) not marked as erasable or not-erasable, the second address page (pag2) containing only address entries not marked as erasable; erasing the first address page (pag1), rendering it ready to be written; writing the content of the second address page into the first and release the second address page for future writings.