Optimizing Distributed AI Workloads in Cloud Environments: A Hybrid Scheduling and Resource Allocation Approach
Current Senior Software Engineer.
Research Article
World Journal of Advanced Research and Reviews, 2024, 23(01), 3137-3149
Publication history:
Received on 27 May 2024; revised on 13 July 2024; accepted on 16 July 2024
Abstract:
Improving performance, scalability and cost efficiency of distributed AI workloads in cloud environment is impossible without optimization. Currently, traditional scheduling and resource allocation methods have shown that they are not able to meet the needs of resources and dynamic characteristics of the AI application. The research discussed in this study aims to identify hybrid scheduling techniques that incorporate static and dynamic strategies, heuristic based techniques, and AI based approaches, for example, reinforcement learning, to optimize work distribution. Furthermore, mechanisms of AI enhanced resource provisioning and cost aware scheduling are explored to enhance the efficiency and cut down the operational costs. The superiority of hybrid AI driven scheduling over conventional approaches is highlighted by the performance comparisons of execution time, energy consumption, scalability, etc. Future research work on edge cloud integration, self-adaptive AI algorithm and quantum computing for AI workload optimization are presented.
Keywords:
Distributed AI; Cloud Computing; Hybrid Scheduling; Resource Allocation; Reinforcement Learning; Workload Optimization
Full text article in PDF:
Copyright information:
Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0