Current opinions on memristor-accelerated machine learning hardware

IF 12.2 2区 材料科学 Q1 MATERIALS SCIENCE, MULTIDISCIPLINARY
Mingrui Jiang , Yichun Xu , Zefan Li , Can Li
{"title":"Current opinions on memristor-accelerated machine learning hardware","authors":"Mingrui Jiang ,&nbsp;Yichun Xu ,&nbsp;Zefan Li ,&nbsp;Can Li","doi":"10.1016/j.cossms.2025.101226","DOIUrl":null,"url":null,"abstract":"<div><div>The unprecedented advancement of artificial intelligence has placed immense demands on computing hardware, but traditional silicon-based semiconductor technologies are approaching their physical and economic limit, prompting the exploration of novel computing paradigms. Memristor offers a promising solution, enabling in-memory analog computation and massive parallelism, which leads to low latency and power consumption. This manuscript reviews the current status of memristor-based machine learning accelerators, highlighting the milestones achieved in developing prototype chips, that not only accelerate neural networks inference but also tackle other machine learning tasks. More importantly, it discusses our opinion on current key challenges that remain in this field, such as device variation, the need for efficient peripheral circuitry, and systematic co-design and optimization. We also share our perspective on potential future directions, some of which address existing challenges while others explore untouched territories. By addressing these challenges through interdisciplinary efforts spanning device engineering, circuit design, and systems architecture, memristor-based accelerators could significantly advance the capabilities of AI hardware, particularly for edge applications where power efficiency is paramount.</div></div>","PeriodicalId":295,"journal":{"name":"Current Opinion in Solid State & Materials Science","volume":"37 ","pages":"Article 101226"},"PeriodicalIF":12.2000,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Current Opinion in Solid State & Materials Science","FirstCategoryId":"88","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1359028625000130","RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0

Abstract

The unprecedented advancement of artificial intelligence has placed immense demands on computing hardware, but traditional silicon-based semiconductor technologies are approaching their physical and economic limit, prompting the exploration of novel computing paradigms. Memristor offers a promising solution, enabling in-memory analog computation and massive parallelism, which leads to low latency and power consumption. This manuscript reviews the current status of memristor-based machine learning accelerators, highlighting the milestones achieved in developing prototype chips, that not only accelerate neural networks inference but also tackle other machine learning tasks. More importantly, it discusses our opinion on current key challenges that remain in this field, such as device variation, the need for efficient peripheral circuitry, and systematic co-design and optimization. We also share our perspective on potential future directions, some of which address existing challenges while others explore untouched territories. By addressing these challenges through interdisciplinary efforts spanning device engineering, circuit design, and systems architecture, memristor-based accelerators could significantly advance the capabilities of AI hardware, particularly for edge applications where power efficiency is paramount.
当前对忆阻器加速机器学习硬件的看法
人工智能的空前进步对计算硬件提出了巨大的要求,但传统的硅基半导体技术正在接近其物理和经济极限,促使人们探索新的计算范式。忆阻器提供了一个很有前途的解决方案,实现了内存模拟计算和大规模并行性,从而降低了延迟和功耗。本文回顾了基于忆阻器的机器学习加速器的现状,强调了在开发原型芯片方面取得的里程碑,这些芯片不仅加速了神经网络推理,而且还解决了其他机器学习任务。更重要的是,它讨论了我们对当前该领域仍然存在的关键挑战的看法,例如器件变化,对高效外围电路的需求,以及系统的协同设计和优化。我们还分享了我们对潜在未来方向的看法,其中一些解决了现有的挑战,而另一些则探索了尚未触及的领域。通过跨越器件工程、电路设计和系统架构的跨学科努力来解决这些挑战,基于忆阻器的加速器可以显著提高人工智能硬件的能力,特别是对于功率效率至关重要的边缘应用。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Current Opinion in Solid State & Materials Science
Current Opinion in Solid State & Materials Science 工程技术-材料科学:综合
CiteScore
21.10
自引率
3.60%
发文量
41
审稿时长
47 days
期刊介绍: Title: Current Opinion in Solid State & Materials Science Journal Overview: Aims to provide a snapshot of the latest research and advances in materials science Publishes six issues per year, each containing reviews covering exciting and developing areas of materials science Each issue comprises 2-3 sections of reviews commissioned by international researchers who are experts in their fields Provides materials scientists with the opportunity to stay informed about current developments in their own and related areas of research Promotes cross-fertilization of ideas across an increasingly interdisciplinary field
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信