Flash Memory Compaction Bitmap for Write Performance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing flash memory compaction algorithms are inefficient, requiring multiple sense operations and data transfers, leading to increased processing cycles and slowed write performance due to inefficient handling of valid and invalid data fragments.
Innovation Solution
A compaction bitmap is used to maintain a one-to-one correspondence with fragments in a source block, indicating valid or invalid data, which is populated in RAM during the compaction process to optimize the transfer of only valid data to a destination block, reducing the number of sense operations and data transfers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a baseline method proceeds fragment by fragment sensing control pages to decipher valid or invalid state, then complete data validation is achieved, but multiple sense operations and data transfers are required increasing processing cycles
Solution Approach 1:
The patent applies preliminary action by pre-populating a compaction bitmap in RAM during the compaction process before actual data transfer. The bitmap is populated by sensing control pages and marking fragment validity states in advance, allowing the system to skip redundant sensing operations during the actual data copy phase. This preliminary preparation reduces the number of sense operations and data transfers required during execution, thereby improving write performance while maintaining complete data validation accuracy.
2Measurement precision
If multiple sense operations are performed on control pages to determine fragment validity, then accurate identification of valid data is achieved, but processing time increases
Solution Approach 1:
The patent uses copying by creating a compaction bitmap that stores copies of fragment validity states in RAM. Instead of repeatedly sensing control pages to determine fragment validity, the system copies the validity information into the bitmap structure during a preliminary phase. This copied information is then referenced during data transfer without requiring additional sense operations, thereby maintaining accurate fragment validity detection while significantly reducing processing time and the number of sense operations required.
3Reliability
If all fragments are processed sequentially to determine validity, then complete compaction is achieved, but the number of data transfers increases
Solution Approach 1:
The patent applies preliminary action by populating the compaction bitmap with fragment validity information before the actual data transfer phase. During this preliminary phase, control pages are sensed and the bitmap is updated to reflect which fragments are valid. Once the bitmap is populated, the data transfer phase only copies valid fragments indicated by the bitmap, avoiding unnecessary transfers of invalid data. This approach ensures complete compaction while minimizing the number of data transfers required.
Solution Approach 2:
The patent extracts only the essential validity information from control pages and stores it in the compaction bitmap. Instead of processing and transferring all fragment data sequentially to determine validity, the system extracts just the validity state of each fragment and stores it in the bitmap. This extracted information is then used to guide the selective copying of only valid fragments, thereby achieving complete compaction while reducing the number of data transfers by eliminating invalid data from the transfer process.
Data Source
AI summary
A memory system may include an optimized data compaction algorithm. The compaction may include transferring only valid data from a source block to a destination block. A compaction bitmap that is maintained in random access memory (“RAM”) may be populated during the compaction process. The populated bitmap may be used to copy valid fragments to the destination block.