Host-Controlled Garbage Collection for Fragmented Nonvolatile Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage devices with nonvolatile memory face challenges in efficiently managing data fragmentation and optimizing data storage operations, particularly in systems using different write modes for varying bit storage capacities, leading to inefficiencies in data handling and resource allocation.
Innovation Solution
An information processing system with a host and a storage device that utilizes a nonvolatile memory and a controller to manage data storage by employing different write modes for memory cells, including a swap area for data not subject to garbage collection, and implements a controller that manages data relationships through tags, allowing for efficient garbage collection processes with progress tracking and host communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If garbage collection is executed to improve data fragmentation, then data management efficiency is improved, but processing time and system resources are consumed
Solution Approach 1:
The host executes a garbage collection analysis command before the actual garbage collection operation to pre-assess the state of storage areas, identify valid data locations, and plan the relocation strategy. This preliminary action allows the system to prepare necessary resources and minimize disruption during the actual garbage collection execution, thereby reducing processing time while maintaining data management efficiency.
Solution Approach 2:
The system implements a feedback mechanism where the host transmits commands to the storage device, receives progress information about valid data identification and relocation status, and adjusts the garbage collection process accordingly. This feedback loop enables optimized resource allocation and real-time monitoring, balancing the trade-off between improving data management efficiency and minimizing processing time consumption.
2Quantity of substance
If different write modes are used for varying bit storage capacities, then storage density is improved, but data handling complexity increases
Solution Approach 1:
The storage device divides its storage areas into multiple segments that can be independently managed with different write modes (e.g., SLC, MLC, TLC, QLC). Each segment can be optimized for specific bit storage capacities based on performance and durability requirements. This segmentation allows the system to achieve high storage density while managing complexity through modular organization and independent control of each segment.
Solution Approach 2:
Different regions or blocks within the storage device are assigned different write modes and characteristics tailored to specific data types or access patterns. For example, frequently accessed data may be stored in faster SLC areas while less critical data uses higher-density QLC areas. This local quality approach optimizes overall system performance and simplifies data handling by matching storage characteristics to data requirements rather than using a uniform approach.
Data Source
AI summary
According to one embodiment, a nonvolatile memory in a storage device includes a swap area storing data that is not selected as an execution target of garbage collection (GC) triggered by a controller in the storage device, is written in a first write mode, and corresponds to a first tag in a write command received from a host. The host transmits, to the storage device, a first command to execute GC. The controller starts the GC in response to the first command received. The controller records progress of the GC in accordance with the GC in progress. The host transmits, to the storage device, a second command to confirm the progress. The controller notifies the host of the progress in response to the second command received.