Unified Data Multi-Operation Graph for Storage Accelerators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hardware-accelerated software-defined storage systems require multiple separate interactions with hardware accelerator devices for performing operations like compression, encryption, and data movement, leading to increased costs and performance issues due to vendor-specific integrations and sequential operation calls.
Innovation Solution
An Information Handling System (IHS) with a data multi-operation management engine that receives a data multi-operation graph, assigns operations to multiple accelerator systems, configures them to perform operations in sequence, and reports completion status, allowing for efficient execution of complex data transformations with reduced overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple separate operation calls are made to hardware accelerator devices, then each operation can be performed with dedicated hardware acceleration, but the number of interactions increases and performance decreases
Solution Approach 1:
The patent combines multiple separate hardware accelerator operations into a single unified operation. The storage application makes one call to the hardware accelerator device to perform multiple operations (compression, encryption, checksum, hash, data movement, RAID, XOR, erasure coding, text search, machine learning inference) simultaneously, rather than making separate calls for each operation. This merging reduces the number of interactions and improves system throughput while maintaining dedicated hardware acceleration for each operation type.
2Reliability
If vendor-specific hardware accelerator interfaces are used, then hardware acceleration is achieved, but costs increase due to custom integrations
Solution Approach 1:
The patent implements a universal interface that allows the storage application to invoke multiple different hardware accelerator operations through a single standardized call mechanism. Instead of requiring separate vendor-specific interfaces for compression, encryption, checksum, hash, data movement, RAID, XOR, erasure coding, text search, and machine learning inference operations, the system provides a unified interface that handles all these operations. This universality reduces integration costs and simplifies the storage application while maintaining hardware acceleration capabilities.
3Ease of operation
If sequential operation calls are made to hardware accelerator devices, then each operation can be controlled independently, but the total execution time increases
Solution Approach 1:
The patent enables continuous execution of multiple hardware accelerator operations within a single operation call. Instead of sequential execution where each operation waits for the previous one to complete, the system performs compression, encryption, checksum, hash, data movement, RAID, XOR, erasure coding, text search, and machine learning inference operations continuously and simultaneously. This maintains independent control of each operation while eliminating the time loss associated with sequential execution.
Data Source
AI summary
A graph-based data multi-operation system includes a data multi-operation management subsystem coupled to an application and accelerator subsystems. The data multi-operation management subsystem receives a data multi-operation graph from the application that identifies first data and defines operations for performance on the first data to transform the first data into second data. The data multi-operation management subsystem assigns each of the operations to at least one of the accelerator systems, and configures the accelerator subsystems to perform the operations in a sequence that transforms the first data into the second data, When the data multi-operation management subsystem determine a completion status for the performance of the operations by the accelerator subsystems, it transmits a completion status communication to the application that indicates the completion status of the performance of the plurality of operations by the plurality of accelerator subsystems.