Scholarlink

This research explores optimizing energy use in AI applications on edge devices like Raspberry Pi. We developed two models to predict resource usage and power consumption of AI algorithms, considering factors like CPU and memory use, algorithm speed, dataset size, and types. By using regression-based methods, we quantified the impact of these factors on energy consumption. The models provide developers with practical tools to evaluate and optimize the energy efficiency of AI deployments on edge servers without sacrificing performance.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Energy-Efficiency Modeling for AI Applications on Edge Computing

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Vamsi Krishna Bhagavathula,
  • Xian Gao,
  • Yi Zhou,
  • Rania Hodhod,
  • Lixin Wang

摘要

This research explores optimizing energy use in AI applications on edge devices like Raspberry Pi. We developed two models to predict resource usage and power consumption of AI algorithms, considering factors like CPU and memory use, algorithm speed, dataset size, and types. By using regression-based methods, we quantified the impact of these factors on energy consumption. The models provide developers with practical tools to evaluate and optimize the energy efficiency of AI deployments on edge servers without sacrificing performance.