| Title |
Performance Evaluation of Edge Computing-based AI Accelerator for Intelligent Battlefield Recognition |
| Authors |
손기영(Kee Young Son) ; 김성국(Seong Guk Kim) ; 윤지원(Ji Won Yoon) ; 진소연(So Yeon Jin) ; 황재민(Jae Min Hwang) ; 김상민(Sang Min Kim) ; 이우신(Woo Sin Lee) ; 김정호(Joung Ho Kim) |
| DOI |
https://doi.org/10.5573/ieie.2022.59.7.41 |
| Keywords |
AI accelerator; Low-power; Military vehicle; Miniaturization; Performance |
| Abstract |
In this paper, the performance and power consumption were evaluated depending on the system configuration of edge computing-based AI accelerator for intelligent battlefield recognition. AI accelerators in military vehicles for intelligent battlefield recognition require high performance, low power, and small form factor. For this requirement, performance and power consumption evaluation of various AI-accelerator configurations with changes in the number of computing unit and memory were conducted and analyzed. The result shows the suitable accelerator system configuration for the requirements of the AI-accelerator in military vehicles for intelligent battlefield recognition. |