Scholarlink

Modern System-on-a-Chip (SoC) designs require very sophisticated power delivery mechanisms which can maintain the load and line regulation of the system with highly nonlinear and variable load demands while maintaining high efficiency throughout the range of output power. This paper proposes a novel approach for a six-phase buck converter design which is controlled by an artificial neural network which can be tailored for highly nonlinear loads. The SoC load is approximated as a variable resistance reacting to the voltage requested by the system as per typical SoC behaviors under loading. The proposed system is modeled in SIMULINK and the simulation results are shown.

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 Neural Network Controlled Multiphase Synchronous Buck Converter for Implementing a Six-Phase Power Delivery Scheme for SoC

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Kumari Pallavi,
  • Soumyakanti Roy,
  • Sushil Kumar Gupta

摘要

Modern System-on-a-Chip (SoC) designs require very sophisticated power delivery mechanisms which can maintain the load and line regulation of the system with highly nonlinear and variable load demands while maintaining high efficiency throughout the range of output power. This paper proposes a novel approach for a six-phase buck converter design which is controlled by an artificial neural network which can be tailored for highly nonlinear loads. The SoC load is approximated as a variable resistance reacting to the voltage requested by the system as per typical SoC behaviors under loading. The proposed system is modeled in SIMULINK and the simulation results are shown.