Scholarlink

In the fast-evolving field of deep learning and convolutional neural networks (CNNs), optimizing power efficiency is of paramount importance. This paper presents an innovative approach to enhance power control within Tiny YOLO CNNs by leveraging the capabilities of Wallace Tree Multiplier (WTM) and Brent Kung Adder (BKA). These hardware components are intricately integrated into the architecture, strategically designed to optimize computational efficiency, while simultaneously minimizing power consumption. Deep learning applications, like image recognition and object detection, often entail high computational demands, resulting in substantial power requirements. Our approach addresses this challenge by intricately weaving the WTM and BKA into the Tiny YOLO framework, yielding significant reductions in power consumption without compromising computational performance. Empirical evaluations affirm the effectiveness of our methodology, underscoring its potential to revolutionize power management in Tiny YOLO CNNs and enhance the sustainability and practicality of deep learning applications across various domains. This research not only addresses the pressing need for power efficiency in deep learning but also demonstrates the feasibility of achieving substantial energy savings through a strategic amalgamation of hardware design and deep learning architecture, with implications extending beyond Tiny YOLO CNNs to influence broader power management in deep learning applications.

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

A Novel Approach for Power Efficiency with Wallace Tree Multiplier and Brent Kung Adder

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • S. Kanagamalliga,
  • K. Monika,
  • P. Monika,
  • R. Latha,
  • S. Rajalingam

摘要

In the fast-evolving field of deep learning and convolutional neural networks (CNNs), optimizing power efficiency is of paramount importance. This paper presents an innovative approach to enhance power control within Tiny YOLO CNNs by leveraging the capabilities of Wallace Tree Multiplier (WTM) and Brent Kung Adder (BKA). These hardware components are intricately integrated into the architecture, strategically designed to optimize computational efficiency, while simultaneously minimizing power consumption. Deep learning applications, like image recognition and object detection, often entail high computational demands, resulting in substantial power requirements. Our approach addresses this challenge by intricately weaving the WTM and BKA into the Tiny YOLO framework, yielding significant reductions in power consumption without compromising computational performance. Empirical evaluations affirm the effectiveness of our methodology, underscoring its potential to revolutionize power management in Tiny YOLO CNNs and enhance the sustainability and practicality of deep learning applications across various domains. This research not only addresses the pressing need for power efficiency in deep learning but also demonstrates the feasibility of achieving substantial energy savings through a strategic amalgamation of hardware design and deep learning architecture, with implications extending beyond Tiny YOLO CNNs to influence broader power management in deep learning applications.