Scholarlink

<p>Optical neural networks (ONNs) leverage the parallelism and low-energy consumption of photonic signal processing to overcome the limitations of traditional electronic computing. Optics inherently enables fan-in and fan-out without the Resistor-Capacitor (RC) and Inductor-Capacitor (LC) delays of electrical interconnects. However, for single-mode photonic integrated circuits, reciprocity constraints introduce unavoidable loss during beam combining, hindering large-scale on-chip photonic fan-in. To overcome this challenge, we provide a photonic lossless mode-division fan-in solution for the convolution accelerators. Using inverse design, we developed a compact multimode photonic convolution accelerator (0.42 mm<sup>2</sup>) with ±15 nm fabrication tolerance and 35 nm optical bandwidth, enabling parallel computation across mode and wavelength dimensions. Experimental results in the C-band confirm a 6–7 bit convolution precision, leading to classification accuracies of 95.2% on MNIST and 87.9% on Fashion-MNIST. Moreover, the device offers a theoretical computational density of 125.14 TOPS/mm<sup>2</sup>, underscoring its potential for scalable and energy-efficient photonic computing accelerators.</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

Highly efficient photonic convolver via lossless mode-division fan-in

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月13日

查看全文
  • Shangsen Sun,
  • Shiji Zhang,
  • Bo Wu,
  • Shan Jiang,
  • Baiheng Zhao,
  • Hailong Zhou,
  • Jianji Dong,
  • Xinliang Zhang

摘要

Optical neural networks (ONNs) leverage the parallelism and low-energy consumption of photonic signal processing to overcome the limitations of traditional electronic computing. Optics inherently enables fan-in and fan-out without the Resistor-Capacitor (RC) and Inductor-Capacitor (LC) delays of electrical interconnects. However, for single-mode photonic integrated circuits, reciprocity constraints introduce unavoidable loss during beam combining, hindering large-scale on-chip photonic fan-in. To overcome this challenge, we provide a photonic lossless mode-division fan-in solution for the convolution accelerators. Using inverse design, we developed a compact multimode photonic convolution accelerator (0.42 mm2) with ±15 nm fabrication tolerance and 35 nm optical bandwidth, enabling parallel computation across mode and wavelength dimensions. Experimental results in the C-band confirm a 6–7 bit convolution precision, leading to classification accuracies of 95.2% on MNIST and 87.9% on Fashion-MNIST. Moreover, the device offers a theoretical computational density of 125.14 TOPS/mm2, underscoring its potential for scalable and energy-efficient photonic computing accelerators.