Memory Refresh Scheduling for Evenly Distributed IO Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing capacity of memory storage devices leads to increased input output (IO) latency due to refresh operations, which negatively impacts user experience.
Innovation Solution
A memory management method that evenly distributes IO latency by executing refresh operations in stages based on the current state and category of physical blocks to be refreshed, using a physical block management table to prioritize and sequence write operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If refresh operations are performed on multiple physical blocks after abnormal power loss, then data reliability is improved, but input output latency time increases
Solution Approach 1:
The refresh operations are segmented and integrated into the normal write operation pipeline. Instead of performing all refresh operations separately before user writes, the system divides refresh tasks into multiple stages that interleave with cache write operations, distributing the refresh burden across multiple I/O cycles rather than concentrating it in one location.
Solution Approach 2:
The patent merges refresh operations with normal user write operations. The controller combines the refresh task (writing refreshed data from physical blocks to cache) with the user's write task (writing user data to cache), so both operations share the same I/O pipeline and resources, reducing total latency.
2Loss of time
If refresh operations are performed before cache is fully written, then IO latency is reduced, but operation complexity increases
Solution Approach 1:
The system dynamically adjusts the execution order and prioritization of refresh operations based on cache status and operation urgency. The controller determines whether to perform refresh operations before, during, or after user writes based on real-time conditions, and assigns different priorities to different physical blocks needing refresh, making the operation schedule flexible rather than fixed.
Solution Approach 2:
The controller uses feedback from the cache write status and refresh operation progress to dynamically adjust the scheduling of subsequent operations. The physical block management table tracks which blocks need refreshing and their priority levels, allowing the controller to make informed decisions about operation sequencing based on current system state.
Data Source
AI summary
A memory management method, a memory storage device, and a memory control circuit unit are provided. The memory management method includes: creating a physical block management table for managing physical blocks; selecting at least one physical block to be refreshed based on the physical block management table when power is restored after an abnormal power loss; in response to a number of the at least one physical block to be refreshed being greater than 1, determining at least one current state and at least one category of the at least one physical block to be refreshed; and executing at least one second write operation based on the current state and the category together when executing a first write operation. The first write operation includes writing user data into a cache. The second write operation includes writing data stored in a physical block to be refreshed into a first physical block.