SSD Zone-Based Compression for Hot and Cold Data Writes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Solid state drives (SSDs) lack efficient data classification, leading to reduced compression efficiency and increased write amplification factor due to random storage of hot and cold data, resulting in higher garbage collection frequencies.

Innovation Solution

A storage system that sets multiple zones in non-volatile memory based on data access attributes, compressing and storing data according to zone-specific methods to differentiate between hot, normal, and cold data, thereby optimizing storage efficiency and reducing garbage collection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If data is randomly stored without classification, then storage simplicity is maintained, but compression efficiency is reduced and write amplification factor increases

Engineering Contradiction:
Improvestorage simplicityVSAvoidcompression efficiency
Core Design Contradiction:
Ease of manufactureVSProductivity

Solution Approach 1:

The storage medium is divided into multiple zones (first zone, second zone, third zone) based on data access patterns. Hot data is stored in the first zone, normal data in the second zone, and cold data in the third zone. This segmentation allows different compression strategies to be applied to different zones, improving overall compression efficiency while maintaining manageable storage organization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different compression ratios are applied to different zones based on their specific characteristics. The first zone (hot data) uses a first compression ratio, the second zone (normal data) uses a second compression ratio, and the third zone (cold data) uses a third compression ratio. This local quality approach optimizes compression efficiency for each zone while balancing write amplification factors.

Inventive Principle:
Principle #3Local quality

2Device complexity

If data is randomly stored without classification, then storage structure simplicity is maintained, but garbage collection frequency increases

Engineering Contradiction:
Improvestorage structure simplicityVSAvoidgarbage collection frequency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The storage medium is segmented into multiple zones based on data access patterns. Hot data is stored in the first zone, normal data in the second zone, and cold data in the third zone. This segmentation isolates frequently accessed data from less accessed data, reducing the impact of garbage collection operations on overall system performance and reducing garbage collection frequency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Data is pre-classified and stored in appropriate zones based on access patterns before garbage collection is needed. By proactively organizing data into hot, normal, and cold zones, the system prepares for future garbage collection operations, reducing the frequency and impact of these operations on system productivity.

Inventive Principle:
Principle #10Preliminary action

3Device complexity

If uniform compression is applied to all data, then processing simplicity is maintained, but storage efficiency is reduced

Engineering Contradiction:
Improveprocessing simplicityVSAvoidstorage efficiency
Core Design Contradiction:
Device complexityVSQuantity of substance

Solution Approach 1:

Different compression ratios are applied to different zones based on their specific characteristics. The first zone (hot data) uses a first compression ratio, the second zone (normal data) uses a second compression ratio, and the third zone (cold data) uses a third compression ratio. This local quality approach optimizes storage efficiency for each zone while maintaining reasonable processing complexity through automated zone identification and compression ratio selection.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The compression ratio is dynamically adjusted based on the zone and data characteristics. Rather than using a static uniform compression ratio, the system dynamically selects appropriate compression ratios for different zones, improving overall storage efficiency while the processing circuitry automatically manages the complexity of determining and applying the appropriate compression ratio for each data block.

Inventive Principle:
Principle #15Dynamics

4Device complexity

If no data classification is performed, then access time simplicity is maintained, but input/output latency increases

Engineering Contradiction:
Improveaccess time simplicityVSAvoidinput/output latency
Core Design Contradiction:
Device complexityVSLoss of time

Solution Approach 1:

The storage medium is segmented into multiple zones based on data access patterns. Hot data is stored in the first zone, normal data in the second zone, and cold data in the third zone. This segmentation improves input/output latency by locating frequently accessed hot data in a dedicated zone that can be quickly accessed, while the processing circuitry automatically identifies and manages zones based on logical addresses, maintaining reasonable access time simplicity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Data is pre-classified and stored in appropriate zones based on access patterns before access is needed. By proactively organizing data into hot, normal, and cold zones, the system prepares for future access operations, reducing input/output latency when data is actually accessed while the automated zone management keeps access time simplicity reasonable.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3869343B1Storage device and operating method thereof
Publication Date: 2023.11.15 SAMSUNG ELECTRONICS CO LTD
  • EP3869343B1 patent drawingFigure 1
  • EP3869343B1 patent drawingFigure 2
  • EP3869343B1 patent drawingFigure 3

AI summary

A storage device may include a non-volatile memory including a plurality of zones, the non-volatile memory configured to sequentially store data in at least one of the plurality of zones, and a processing circuitry configured to, receive a first write command and first data from a host, the first write command including a first logical address, identify a first zone of the plurality of zones based on the first logical address, compress the first data based on compression settings corresponding to the first zone, and write the compressed first data to the first zone.