AI Management Platform with Embedded Edge Cluster

Download PDF
Open Webpage

Seung Woo Kum, Miseon Yu, Youngkee Kim, Jaewon Moon, Silvio Cretti

Published: 2021, Last Modified: 06 Oct 2025ICCE 2021EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: The area of AI application continues to grow from resource-rich environment such as cloud to resource-constraint environment such as embedded devices. However, management of AI on the embedded device is still a challenging task due to different environment other that resource-rich devices. This paper proposes an AI management platform for a cluster which consists of edge devices and cloud resources. The proposed system provides efficient methods to identify different hardware acceleration resources and resource management for a specific AI application. The implementation of the proposed platform on a cluster which consists of cloud VMs and embedded devices are described presented.