Sankar,, Thambireddy and Venkata Ramana, Reddy Bussu and Balamuralikrishnan, Anbalagan (2023) AI-Optimized Hyperscale Data Centers: Meeting the Rising Demands of Generative AI Workloads. International Journal of Trend in Scientific Research and Development, 7 (1). pp. 1504-1514. ISSN 2456-6470

Text ijtsrd52785.pdf Download (1MB)

Abstract

The sheer increase in generative artificial intelligence (AI) applications, including large language models and creative generative innovations, has stressed data centre infrastructure regarding computation. Generative AI is known to have unique needs, especially regarding data centre configurations and demands a generally efficient but not necessarily optimized data centre solution built to handle generalized workloads. In meeting these growing demands, this paper discusses the developments where hyper-scale data centres are being redesigned and optimized using AI to enable such requirements. It covers architecture, hardware accelerators, thermal management, energy consumption and orchestration systems that are now part and parcel of generative AI in scale. Moreover, the paper outlines the significant challenges, such as the energy-intensive aspect, latency, and sustainability issues, and provides examples of leading companies that introduce innovative solutions. It is summarized with proactive guiding recommendations towards establishing resilience, efficiency, and scalable data centre infrastructures compatible with the next generation of the evolution of AI.

Item Type:	Article
Subjects:	P Language and Literature > PM Hyperborean, Indian, and Artificial languages
Divisions:	Postgraduate > Master's of Islamic Education
Depositing User:	Journal Editor
Date Deposited:	04 Sep 2025 09:54
Last Modified:	04 Sep 2025 09:54
URI:	http://eprints.umsida.ac.id/id/eprint/16332

Actions (login required)

View Item