Scholarlink

<p>Layered metal dichalcogenides (LMDs) neuromorphic memristor devices offer a promising alternative to conventional von Neumann architectures, addressing speed and energy efficiency constraints. However, challenges remain in controlling resistive switching and operating voltage in crystalline LMD memristors due to environmental stabilization issues, which hinder neural network hardware development. Herein, we introduce an optimization method for memristor operation by controlling oxidation through ozone treatment, creating a SnO<sub><i>x</i></sub>/SnS<sub>2</sub> resistive layer. These optimized memristors demonstrate low switching voltages (∼1 V), rapid switching speeds (∼20 ns), high switching ratios (10<sup>2</sup>), and the ability to emulate synaptic weight plasticity. Cross-sectional transmission electron microscopy and energy-dispersive X-ray spectroscopy identified defects and Ti conductive filaments in the resistive switching layer, contributing to uniform switching and minimized operating variation. The device achieved 90% accuracy in MNIST handwritten recognition, and hardware-based image convolution was successfully implemented, showcasing the potential of SnO<sub><i>x</i></sub>/SnS<sub>2</sub> memristors for neuromorphic applications.</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

Enhancing memristor performance with 2D SnOx/SnS2 heterostructure for neuromorphic computing

  • 期刊论文

  • 发表日期：

    2025年1月2日

查看全文
  • Yangwu Wu,
  • Sifan Li,
  • Yun Ji,
  • Zhengjin Weng,
  • Houying Xing,
  • Lester Arauz,
  • Travis Hu,
  • Jinhua Hong,
  • Kah-Wee Ang,
  • Song Liu

摘要

Layered metal dichalcogenides (LMDs) neuromorphic memristor devices offer a promising alternative to conventional von Neumann architectures, addressing speed and energy efficiency constraints. However, challenges remain in controlling resistive switching and operating voltage in crystalline LMD memristors due to environmental stabilization issues, which hinder neural network hardware development. Herein, we introduce an optimization method for memristor operation by controlling oxidation through ozone treatment, creating a SnOx/SnS2 resistive layer. These optimized memristors demonstrate low switching voltages (∼1 V), rapid switching speeds (∼20 ns), high switching ratios (102), and the ability to emulate synaptic weight plasticity. Cross-sectional transmission electron microscopy and energy-dispersive X-ray spectroscopy identified defects and Ti conductive filaments in the resistive switching layer, contributing to uniform switching and minimized operating variation. The device achieved 90% accuracy in MNIST handwritten recognition, and hardware-based image convolution was successfully implemented, showcasing the potential of SnOx/SnS2 memristors for neuromorphic applications.