In-Storage Computing Controller for Data Extraction Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data server systems face inefficiencies in data extraction and traffic management due to the need for processors to handle large amounts of data, leading to high computing power requirements and network traffic.

Innovation Solution

A storage device with a non-volatile memory and controller, capable of receiving and processing data processing codes generated by a host, performing in-storage computing operations, and transmitting results back to the host, thereby reducing the load on the host and network traffic.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data is transmitted from database to server and processed by server processor, then data extraction can be performed, but data traffic between server and database increases and computing power requirement increases

Engineering Contradiction:

Improvedata extraction efficiencyVSAvoidnetwork traffic and computing power consumption

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The patent extracts the data processing function from the server and relocates it to the storage device. The storage device receives data from the database and performs extraction operations locally, then transmits only the extracted results to the server. This extraction of processing capability reduces network traffic and server computing power requirements while maintaining data extraction efficiency.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The storage device acts as an intermediary between the database and the server. It receives data from the database, performs processing operations, and sends results to the server. This intermediary role reduces the direct data traffic burden on the server and enables more efficient data extraction without requiring the server to handle all processing tasks.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If processor extracts desired data from big data in main memory, then data extraction is performed, but data traffic between server and database is not decreased

Engineering Contradiction:

Improvedata extraction capabilityVSAvoiddata traffic volume

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The storage device performs data extraction operations before the data reaches the server. By conducting preliminary processing at the storage device, the system reduces the volume of data that needs to be transmitted to the server, thereby decreasing network traffic while maintaining the server's data extraction capability.

Inventive Principle:

Principle #10Preliminary action

3Loss of energy

If storage device performs in-storage computing operations, then network traffic and computing power requirement are reduced, but device complexity increases

Engineering Contradiction:

Improvenetwork traffic and computing power consumptionVSAvoidstorage device structure

Core Design Contradiction:

Loss of energyVSDevice complexity

Solution Approach 1:

The storage device is designed with multi-functionality, combining storage capabilities with computing functions. The controller includes a CPU with multiple cores that can execute data processing codes, and the device can perform various computing operations directly on stored data. This universal design reduces network traffic and computing power requirements while managing device complexity through integrated functionality.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent implements a nested architecture where the storage device contains an on-chip memory that can overlay data processing codes. The controller includes a CPU with multiple cores, and the system can store processing codes in the non-volatile memory. This nested structure enables in-storage computing while managing complexity through hierarchical organization of components.

Inventive Principle:

Principle #7Nested doll (Nesting)

Data Source

PatentUS9652416B2Storage device for performing in-storage computing operations, method of operation the same, and system including the same

Publication Date: 2017.05.16 SAMSUNG ELECTRONICS CO LTD

AI summary

A storage device performs in-storage computing operation, and includes a non-volatile memory configured to store data and a controller. The controller may include an on-chip memory and may control an operation of the non-volatile memory. The controller receives a data processing code generated by a host, overlays the data processing code on the on-chip memory, processes first data corresponding to the data processing code among the data stored in the non-volatile memory, and transmits the processed first data to the host.