Calculation power configuration method and device based on resource demand prediction, equipment, medium and product

By using resource demand prediction model and BKD tree index in multi-core heterogeneous computing clusters and combining with gray wolf optimization algorithms, the problem of low computing power configuration accuracy in the existing technology is solved, and more efficient resource utilization and computing performance is achieved.

CN119718663AActive Publication Date: 2025-03-28CHINA SOUTHERN POWER GRID COMPANY
4 Cites 1 Cited by

Patent Information

Application Number
CN202411893373.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-20
Publication Date
2025-03-28
Estimated Expiration
2044-12-20

AI Technical Summary

Technical Problem

The accuracy of computing power configuration in the prior art in multi-core heterogeneous computing cluster is low, and other influencing factors other than GPUs are not effectively considered.

Method used

Through a resource demand prediction method, the pre-trained resource demand prediction model is inputted using historical performance data and historical log data to predict the execution time of the task, and a BKD tree index is constructed to determine the local and global computing power configuration. Combined with the Gray Wolf optimization algorithm, optimize computing power configuration to improve accuracy.

Benefits of technology

The computing power configuration accuracy of multi-core heterogeneous computing power clusters is improved, and resource utilization and computing efficiency are improved by more accurately predicting task execution time and optimizing resource utilization.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to a computing power configuration method and device based on resource demand prediction, equipment, a medium and a product, and relates to the technical field of system resource management, and the method comprises the steps: inputting historical performance data and historical log data of a multi-core heterogeneous computing power cluster into a pre-trained resource demand prediction model, obtaining a plurality of predicted tasks and predicted execution time corresponding to each task; constructing a BKD tree index corresponding to each resource dimension according to each computing power node of the multi-core heterogeneous computing power cluster and the resource dimensions corresponding to the plurality of tasks, and representing a corresponding relationship between each resource dimension and each task and a load condition of the computing power node for each resource dimension; determining local target computing power configuration corresponding to each resource dimension according to each computing power node and the BKD tree index; and determining global target computing power configuration for the multi-core heterogeneous computing power cluster according to the predicted execution time corresponding to each task, the local target computing power configuration and a grey wolf optimization algorithm. By adopting the method, the accuracy can be improved.
Need to check novelty before this filing date? Find Prior Art