Scholarlink

<p>Networks-on-Chip (NoCs) address the communication challenges that arise from integrating numerous IP cores onto a single chip. This paper explores the use of encoding algorithms to reduce power dissipation in NoCs. To be effective, the power saved on interconnect links must outweigh the overhead introduced by encoder and decoder modules. To this end, the paper introduces a novel criterion for evaluating the efficiency of low-power encoding in NoCs. Based on this, it proposes the&#xa0;Adaptive Low Power encoding Scheme (ALPS) an adaptive method that selectively bypasses encoding modules at runtime when their power overhead exceeds the benefit of link-level power savings. Experimental results show that ALPS reduces overall NoC power consumption by up to&#xa0;14.6%&#xa0;on average compared to state-of-the-art low-power encoding techniques.</p>

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

ALPS: adaptive low power encoding scheme for networks on chip

  • 期刊论文

  • 发表日期：

    2026年2月11日

查看全文
  • Mehdi Taassori

摘要

Networks-on-Chip (NoCs) address the communication challenges that arise from integrating numerous IP cores onto a single chip. This paper explores the use of encoding algorithms to reduce power dissipation in NoCs. To be effective, the power saved on interconnect links must outweigh the overhead introduced by encoder and decoder modules. To this end, the paper introduces a novel criterion for evaluating the efficiency of low-power encoding in NoCs. Based on this, it proposes the Adaptive Low Power encoding Scheme (ALPS) an adaptive method that selectively bypasses encoding modules at runtime when their power overhead exceeds the benefit of link-level power savings. Experimental results show that ALPS reduces overall NoC power consumption by up to 14.6% on average compared to state-of-the-art low-power encoding techniques.