Scholarlink

<p>For neuromorphic computing, integrating multiply-accumulate operations and nonlinear activation within a single device can reduce latency and power consumption while improving computational efficiency. However, these operations impose conflicting requirements: multiply-accumulate requires highly linear and non-volatile resistance states, while activation requires diverse nonlinear synaptic behaviors. These contrasting demands make integrating both functions in a single device challenging. Here, we construct a multimodal ion-gate transistor using 2D CdPS<sub>3</sub>-Li as dielectric layer and MoS<sub>2</sub> as channel material. The layered structure of CdPS<sub>3</sub>-Li facilitates anisotropic ion transport for Li<sup>+</sup> storage and produces strong ion-electron coupling, resulting in high-linearity and non-volatile resistance states under electrical pulses. Moreover, Cd vacancies in CdPS<sub>3</sub>-Li attract and trap photo-generated holes from MoS<sub>2</sub>, leading to rich nonlinear behavior under light pulses. Therefore, the CdPS<sub>3</sub>-Li transistor can simultaneously perform both operations. The CdPS<sub>3</sub>-Li transistor arrays achieved high accuracy in handwritten digit classification, offering a promising hardware solution for neuromorphic computing.</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

Multimodal ion-gated transistor based on 2D superionic conductor for in-memory computing in deep learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月18日

查看全文
  • Bo Tong,
  • Tianming Du,
  • Jinhong Du,
  • Xin Yu,
  • Jinmeng Tong,
  • Jiajun Xu,
  • Wanying Li,
  • Zhibo Liu,
  • Chen Li,
  • Wencai Ren

摘要

For neuromorphic computing, integrating multiply-accumulate operations and nonlinear activation within a single device can reduce latency and power consumption while improving computational efficiency. However, these operations impose conflicting requirements: multiply-accumulate requires highly linear and non-volatile resistance states, while activation requires diverse nonlinear synaptic behaviors. These contrasting demands make integrating both functions in a single device challenging. Here, we construct a multimodal ion-gate transistor using 2D CdPS3-Li as dielectric layer and MoS2 as channel material. The layered structure of CdPS3-Li facilitates anisotropic ion transport for Li+ storage and produces strong ion-electron coupling, resulting in high-linearity and non-volatile resistance states under electrical pulses. Moreover, Cd vacancies in CdPS3-Li attract and trap photo-generated holes from MoS2, leading to rich nonlinear behavior under light pulses. Therefore, the CdPS3-Li transistor can simultaneously perform both operations. The CdPS3-Li transistor arrays achieved high accuracy in handwritten digit classification, offering a promising hardware solution for neuromorphic computing.