Computational Storage Delimiters for Parallel Host Data Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computational storage devices face challenges in providing parallel execution pipelines due to misalignment of host data units across multiple components, making it difficult to perform computations efficiently.
Innovation Solution
Implementing a system with first and second processors and buffers that utilize delimiters to identify and process host data units across storage sources, ensuring computations are performed on complete units by aligning with delimiter locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If computations are offloaded to CSD and targeted to host data units, then computational efficiency is improved, but parallel execution is difficult due to unknown alignment of host data units across storage components
Solution Approach 1:
The patent introduces delimiters as intermediary markers that are inserted between storage components to indicate boundaries of host data units. These delimiters act as mediators that enable processors to identify complete host data units across multiple storage components without requiring knowledge of alignment, thus enabling parallel execution while maintaining computational efficiency
Solution Approach 2:
The system performs preliminary actions by inserting delimiters into data during the storage process, before computations are executed. This preliminary marking of data boundaries allows subsequent parallel processing to proceed efficiently without requiring complex alignment detection, as the delimiters have already prepared the data structure for easy identification of complete host data units
2Quantity of substance
If host data units are split across multiple storage components, then storage capacity is improved, but alignment information is lost making parallel processing difficult
Solution Approach 1:
Delimiters serve as intermediary markers that are inserted between storage components to indicate boundaries of host data units. These delimiters act as mediators that enable processors to identify complete host data units across multiple storage components without requiring knowledge of alignment, thus enabling parallel execution while maintaining computational efficiency
Solution Approach 2:
The patent changes the parameter representation by transforming implicit alignment information into explicit delimiter markers. Instead of relying on the original alignment parameters that are lost when data is split across storage components, the system introduces a new parameter format (delimiters) that explicitly marks boundaries, making the information accessible and usable for parallel processing
Data Source
AI summary
A storage device includes a first storage source storing first data. The storage device further includes a second storage source storing second data. The second data includes a first portion and a second portion separated by a delimiter. The storage device further includes a first buffer configured to receive the first data. The storage device further includes a second buffer configured to receive the second data. The storage device further includes a first processor associated with the first buffer. The storage device further includes a second processor associated with the second buffer. The second processor is configured to perform a first operation on the second portion of the second data, and wherein the first processor is configured to perform a second operation on the first data and the first portion of the second data based on the delimiter.