Acceleration Device Task Assignment by Computing Resource Needs

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing storage systems inefficiently utilize acceleration devices due to imbalanced load distribution caused by varying computing resource requirements of different tasks, leading to waste of computing resources and suboptimal performance.

Innovation Solution

A method for assigning tasks to acceleration devices based on the specific computing resources required or requested, determining the resource needs of each task queue and selecting the most suitable device for optimal load balancing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If tasks are assigned to acceleration devices without considering computing resource requirements, then assignment simplicity is maintained, but load balancing deteriorates and computing resource waste increases

Engineering Contradiction:

Improvetask assignment simplicityVSAvoidsystem efficiency

Core Design Contradiction:

Ease of operationVSProductivity

Solution Approach 1:

The patent introduces computing resource requirements as a new parameter for task assignment decisions. By evaluating the computing resource requirements of tasks and matching them with the capabilities and current load states of acceleration devices, the system achieves optimized load balancing while maintaining efficient resource utilization, resolving the contradiction between assignment simplicity and system efficiency

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces simple round-robin or random task assignment mechanisms with an intelligent selection mechanism that evaluates computing resource requirements. This substitution transforms the task assignment from a mechanical sequential process to an intelligent decision-making process that considers task characteristics and device states, thereby improving productivity without significantly increasing operational complexity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If tasks are assigned based on computing resource requirements, then load balancing and system efficiency are improved, but assignment complexity increases

Engineering Contradiction:

Improvesystem efficiencyVSAvoidtask assignment complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent performs preliminary evaluation of computing resource requirements for tasks before assignment decisions are made. By pre-assessing the computing resource needs of tasks and the current states of acceleration devices, the system prepares necessary information in advance, which simplifies the actual assignment process while maintaining high efficiency and load balancing

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent enables acceleration devices to self-report their current load states and computing resource availability. This self-service mechanism allows the task assignment system to obtain real-time device information without complex monitoring infrastructure, reducing assignment complexity while achieving optimal load distribution

Inventive Principle:

Principle #25Self-service

3Productivity

If computing resources are not properly balanced across acceleration devices, then device utilization is improved, but resource waste increases

Engineering Contradiction:

Improvedevice utilizationVSAvoidcomputing resource waste

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The patent implements a feedback mechanism where the task assignment system continuously monitors the load states and computing resource consumption of acceleration devices. Based on this feedback information, the system dynamically adjusts task assignment decisions to balance computing resources across devices, preventing both underutilization and waste of computing resources, thereby resolving the contradiction between device utilization and resource waste

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20250321796A1Method, device, and computer program product for assigning tasks to acceleration devices

Publication Date: 2025.10.16 DELL PROD LP

AI summary

The described technology relates to assigning tasks to acceleration devices. For instance, an example method includes determining a target task to be assigned to multiple acceleration devices of a storage system. Here, the acceleration devices are configured to handle various types of tasks. The method further includes determining multiple corresponding computing resources required or requested to perform multiple corresponding task queues of the multiple acceleration devices. The method further includes selecting a target acceleration device from the multiple acceleration devices based on the multiple computing resources. The method further includes assigning the target task to the target acceleration device. Using the described technology, the scheduling of target tasks can be optimized on the multiple acceleration devices and the computing resources can be assigned more reasonably, thereby optimizing the system performance.