SSD Token Control for Uniform Garbage Collection Writes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing storage devices face challenges in maintaining uniform execution of background and foreground operations, leading to irregular quality of service, particularly in high-performance semiconductor-based solid-state drives used in data centers and cloud computing environments.

Innovation Solution

A storage device with a memory controller that generates primary and secondary tokens based on the size of valid pages and consumption amount of victim blocks to manage background and foreground operations, ensuring balanced execution of write operations through token-based management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If the storage device performs background operations (garbage collection) in parallel with foreground operations (host requests), then the storage capacity and data management are improved, but the quality of service uniformity deteriorates

Engineering Contradiction:
Improvestorage capacityVSAvoidquality of service uniformity
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent introduces a token-based intermediary mechanism where tokens are generated based on valid page sizes and consumed during write operations. This token system acts as a mediator that regulates and coordinates both foreground and background operations, ensuring they execute uniformly without compromising service quality while maintaining storage capacity management.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Productivity

If the storage device alternates between background operations and foreground operations, then resource management is improved, but operational efficiency deteriorates

Engineering Contradiction:
Improveoperational efficiencyVSAvoidoperation management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent enables continuous execution of both foreground and background operations simultaneously through the token mechanism. Instead of alternating operations, the system allows concurrent execution where tokens are generated continuously based on valid page sizes and consumed by both types of operations, maintaining high productivity while simplifying operation management.

Inventive Principle:
Principle #20Continuity of useful action

3Reliability

If the storage device performs garbage collection on victim blocks, then data management quality is improved, but the execution uniformity of operations deteriorates

Engineering Contradiction:
Improvedata management qualityVSAvoidoperation execution uniformity
Core Design Contradiction:
ReliabilityVSStability of the object's composition

Solution Approach 1:

The patent applies local quality by generating tokens specifically based on the valid page size of victim blocks targeted for garbage collection. This localized token generation ensures that garbage collection operations receive appropriate resource allocation based on their specific characteristics, maintaining data management quality while ensuring uniform execution across different operation types through the standardized token interface.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS20260072589A1Storage device and operating method thereof
Publication Date: 2026.03.12 SAMSUNG ELECTRONICS CO LTD
  • US20260072589A1 patent drawing
  • US20260072589A1 patent drawing
  • US20260072589A1 patent drawing

AI summary

A storage device including a memory device that includes a plurality of nonvolatile memory blocks; and a memory controller that controls the memory device. The memory controller performs a first write operation based on a request of a host and a second write operation based on garbage collection on the memory device, based on primary tokens and secondary tokens. The memory controller generates the primary tokens based on a size of a valid page of a victim block from among the plurality of nonvolatile memory blocks, and generates the secondary tokens based on a consumption amount of the primary tokens.