View in EDS

A sub–10-millisecond neural dynamical system based on phase-change memristors.

Saved in:
Bibliographic Details
Title: A sub–10-millisecond neural dynamical system based on phase-change memristors.
Authors: Cai, Lei (AUTHOR), Tao, Yaoyu (AUTHOR), Xie, Chenchen (AUTHOR), Yan, Longhao (AUTHOR), Li, Shiqian (AUTHOR), Shen, Ruihong (AUTHOR), Pan, Zelun (AUTHOR), Wang, Xile (AUTHOR), Wang, Bowen (AUTHOR), Shi, Daijing (AUTHOR), Zhu, Yihang (AUTHOR), Zhang, Teng (AUTHOR), Zhu, Yixin (AUTHOR), Li, Xi (AUTHOR), Song, Zhitang (AUTHOR), Huang, Ru (AUTHOR), Yang, Yuchao (AUTHOR)
Source: Science. 7/2/2026, Vol. 393 Issue 6806, p105-112. 8p.
Subjects: Memristors, Dynamical systems, Computer architecture, Real-time computing, Mathematical optimization, Three-dimensional modeling
Abstract: High-fidelity geometry for physical-world modeling demands real-time, dense, and differentiable deformation fields on manifolds. Neural dynamical systems (NDSs) using adaptive stepsize integration with embedded neural networks excel at these tasks but still suffer latency on the order of hundreds of milliseconds. In this work, we report a sub–10-millisecond NDS hardware leveraging the precisely controlled conductance drift of phase-change memristors and their multilevel compute-in-memory capabilities. We fabricated a 40-nanometer NDS chip for the challenging surface reconstruction tasks. Compared with state-of-the-art NDS hardware, our NDS design achieves a latency of 2.12 milliseconds (below 10 milliseconds) for single-iteration NDS computations with an error tolerance of 10−7 and delivers 3.82× to 36.27× faster speed while consuming 11.75× to 24.73× less power. The end-to-end NDS latency through hardware measurements and simulations outperformed graphics processing unit A100 by 50.38× to 478.18×. Editor's summary: One of the challenges of using a phase-change memristor (PCM) for analog computing is its natural conductance drift caused by structural relaxation. However, such drift could serve as a functional computing feature rather than as a limitation for a neural dynamical system (NDS) that synergizes the strengths of neural networks and the power of continuous-time dynamical systems for various complex physical world models. By exploiting two distinct characteristics of PCMs, precision-controlled conductance drift and multilevel conductance, Cai et al. demonstrated a hardware-algorithm codesign that delivered significant speed and power improvements over digital NDS accelerators (see the Perspective by Zheng and Valov). The custom chip enabled sub–10 millisecond low-latency NDS computation, supported complex three-dimensional modeling tasks, and achieved up to 5089 times speedup with much lower power consumption compared with state-of-the-art NDS accelerators. —Yury Suleymanov [ABSTRACT FROM AUTHOR]
Copyright of Science is the property of American Association for the Advancement of Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Psychology and Behavioral Sciences Collection
FullText Text:
  Availability: 0
Header DbId: pbh
DbLabel: Psychology and Behavioral Sciences Collection
An: 195069947
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: A sub–10-millisecond neural dynamical system based on phase-change memristors.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Cai%2C+Lei%22">Cai, Lei</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Tao%2C+Yaoyu%22">Tao, Yaoyu</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Xie%2C+Chenchen%22">Xie, Chenchen</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yan%2C+Longhao%22">Yan, Longhao</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Shiqian%22">Li, Shiqian</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Shen%2C+Ruihong%22">Shen, Ruihong</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Pan%2C+Zelun%22">Pan, Zelun</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Xile%22">Wang, Xile</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Bowen%22">Wang, Bowen</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Shi%2C+Daijing%22">Shi, Daijing</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Zhu%2C+Yihang%22">Zhu, Yihang</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Zhang%2C+Teng%22">Zhang, Teng</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Zhu%2C+Yixin%22">Zhu, Yixin</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Xi%22">Li, Xi</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Song%2C+Zhitang%22">Song, Zhitang</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Huang%2C+Ru%22">Huang, Ru</searchLink> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yang%2C+Yuchao%22">Yang, Yuchao</searchLink> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Science%22">Science</searchLink>. 7/2/2026, Vol. 393 Issue 6806, p105-112. 8p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Memristors%22">Memristors</searchLink><br /><searchLink fieldCode="DE" term="%22Dynamical+systems%22">Dynamical systems</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+architecture%22">Computer architecture</searchLink><br /><searchLink fieldCode="DE" term="%22Real-time+computing%22">Real-time computing</searchLink><br /><searchLink fieldCode="DE" term="%22Mathematical+optimization%22">Mathematical optimization</searchLink><br /><searchLink fieldCode="DE" term="%22Three-dimensional+modeling%22">Three-dimensional modeling</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: High-fidelity geometry for physical-world modeling demands real-time, dense, and differentiable deformation fields on manifolds. Neural dynamical systems (NDSs) using adaptive stepsize integration with embedded neural networks excel at these tasks but still suffer latency on the order of hundreds of milliseconds. In this work, we report a sub–10-millisecond NDS hardware leveraging the precisely controlled conductance drift of phase-change memristors and their multilevel compute-in-memory capabilities. We fabricated a 40-nanometer NDS chip for the challenging surface reconstruction tasks. Compared with state-of-the-art NDS hardware, our NDS design achieves a latency of 2.12 milliseconds (below 10 milliseconds) for single-iteration NDS computations with an error tolerance of 10−7 and delivers 3.82× to 36.27× faster speed while consuming 11.75× to 24.73× less power. The end-to-end NDS latency through hardware measurements and simulations outperformed graphics processing unit A100 by 50.38× to 478.18×. Editor's summary: One of the challenges of using a phase-change memristor (PCM) for analog computing is its natural conductance drift caused by structural relaxation. However, such drift could serve as a functional computing feature rather than as a limitation for a neural dynamical system (NDS) that synergizes the strengths of neural networks and the power of continuous-time dynamical systems for various complex physical world models. By exploiting two distinct characteristics of PCMs, precision-controlled conductance drift and multilevel conductance, Cai et al. demonstrated a hardware-algorithm codesign that delivered significant speed and power improvements over digital NDS accelerators (see the Perspective by Zheng and Valov). The custom chip enabled sub–10 millisecond low-latency NDS computation, supported complex three-dimensional modeling tasks, and achieved up to 5089 times speedup with much lower power consumption compared with state-of-the-art NDS accelerators. —Yury Suleymanov [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Science is the property of American Association for the Advancement of Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=pbh&AN=195069947
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1126/science.aee6277
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 8
        StartPage: 105
    Subjects:
      – SubjectFull: Memristors
        Type: general
      – SubjectFull: Dynamical systems
        Type: general
      – SubjectFull: Computer architecture
        Type: general
      – SubjectFull: Real-time computing
        Type: general
      – SubjectFull: Mathematical optimization
        Type: general
      – SubjectFull: Three-dimensional modeling
        Type: general
    Titles:
      – TitleFull: A sub–10-millisecond neural dynamical system based on phase-change memristors.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Cai, Lei
      – PersonEntity:
          Name:
            NameFull: Tao, Yaoyu
      – PersonEntity:
          Name:
            NameFull: Xie, Chenchen
      – PersonEntity:
          Name:
            NameFull: Yan, Longhao
      – PersonEntity:
          Name:
            NameFull: Li, Shiqian
      – PersonEntity:
          Name:
            NameFull: Shen, Ruihong
      – PersonEntity:
          Name:
            NameFull: Pan, Zelun
      – PersonEntity:
          Name:
            NameFull: Wang, Xile
      – PersonEntity:
          Name:
            NameFull: Wang, Bowen
      – PersonEntity:
          Name:
            NameFull: Shi, Daijing
      – PersonEntity:
          Name:
            NameFull: Zhu, Yihang
      – PersonEntity:
          Name:
            NameFull: Zhang, Teng
      – PersonEntity:
          Name:
            NameFull: Zhu, Yixin
      – PersonEntity:
          Name:
            NameFull: Li, Xi
      – PersonEntity:
          Name:
            NameFull: Song, Zhitang
      – PersonEntity:
          Name:
            NameFull: Huang, Ru
      – PersonEntity:
          Name:
            NameFull: Yang, Yuchao
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 02
              M: 07
              Text: 7/2/2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 00368075
          Numbering:
            – Type: volume
              Value: 393
            – Type: issue
              Value: 6806
          Titles:
            – TitleFull: Science
              Type: main
ResultId 1