NAND Controller Latency via L2P Region Bitmap
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional memory devices face inefficiencies in garbage collection due to the large size of translation tables, leading to high latencies and reduced performance, especially when the complete table does not fit within working memory, causing unnecessary searches and increased wait times during maintenance tasks like garbage collection.
Innovation Solution
Implementing a data structure such as a bitmap to track relevant logical-to-physical (L2P) table regions, allowing only pertinent regions to be loaded into working memory during garbage collection, and modifying the cadence of garbage collection operations to spread latency caused by stale data structure entries across multiple host writes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the complete L2P translation table is loaded into working memory for garbage collection, then search accuracy is improved, but memory capacity requirements increase and latency increases when the table does not fit in working memory
Solution Approach 1:
The L2P translation table is divided into multiple regions, and only the relevant regions are loaded into working memory during garbage collection operations. This segmentation allows the system to work with portions of the table rather than the entire table, reducing memory capacity requirements while maintaining search accuracy for the relevant portions.
Solution Approach 2:
The system pre-identifies and loads only the pertinent L2P table regions into working memory before garbage collection operations begin. This preliminary action ensures that the required data is readily available in fast memory, improving search accuracy while avoiding the need to load the complete table.
2Reliability
If all L2P table regions are searched during garbage collection, then completeness of garbage collection is improved, but operation time increases
Solution Approach 1:
The system extracts and loads only the pertinent L2P table regions into working memory, excluding irrelevant regions from the garbage collection search process. This extraction maintains the completeness of garbage collection for the relevant data while significantly reducing the time required to search through the entire translation table.
3Loss of time
If garbage collection operations are performed frequently to reduce latency, then responsiveness is improved, but system overhead increases
Solution Approach 1:
The system performs partial garbage collection operations by processing only the pertinent L2P table regions that are currently loaded in working memory. This partial action reduces the overhead of each garbage collection operation while maintaining acceptable latency by focusing resources on the most relevant data portions.
Data Source
AI summary
Devices and techniques for controlling NAND operation latency are described herein. A controlled can receive a write request. The controller can then calculate a number of garbage collection operations to perform on a physical block that is closed. Here, the calculation includes adding a logical-to-physical (L2P) region search ratio to a cadence calculation for garbage collection. Garbage collection operations can be performed on the physical block in accordance with the calculated number of garbage collection operations to perform. Then, the controller can perform the write request in response to completing the calculated number of garbage collection operations.