Yiming Sun, Xing Chen, Chao Chen, Baojia Liu, Bingyu Chen, Zhiyuan Zhao, Dahai Wei, Christian H. Back, Wang Kang, Weisheng Zhao, Na Lei
{"title":"多铁氧体异质结构中时空共复用的应变介导储层计算","authors":"Yiming Sun, Xing Chen, Chao Chen, Baojia Liu, Bingyu Chen, Zhiyuan Zhao, Dahai Wei, Christian H. Back, Wang Kang, Weisheng Zhao, Na Lei","doi":"10.1063/5.0221747","DOIUrl":null,"url":null,"abstract":"Physical reservoir computing (PRC), a brain-inspired computing method known for its efficient information processing and low training requirements, has attracted significant attention. The key factor lies in the number of computational nodes within the reservoir for its computational capability. Here, we explore co-multiplexing reservoirs that leverage both temporal and spatial strategies. Temporal multiplexing virtually expands the node count through the use of masking techniques, while spatial multiplexing utilizes multiple physical locations (e.g., Hall bars) to achieve an increase in the number of real nodes. Our experiment employs a strain-mediated reservoir based on multiferroic heterostructures. By applying a single voltage across the PMN-PT substrate (acting as global input) and measuring the output Hall voltages from four Hall bars (real nodes), we achieve significant efficiency gains. This co-multiplexing approach results in a reduction in the normalized root mean square error from 0.5 to 0.23 for a 20-step prediction task of a Mackey–Glass chaotic time series. Furthermore, the single input and four independent outputs lead to a fourfold reduction in energy consumption compared to the strain-mediated PRC with temporal multiplexing solely. This research paves the way for future energy saving PRC implementations utilizing co-multiplexing, promoting a resource-efficient paradigm in reservoir computing.","PeriodicalId":8094,"journal":{"name":"Applied Physics Letters","volume":null,"pages":null},"PeriodicalIF":3.5000,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Strain-mediated reservoir computing with temporal and spatial co-multiplexing in multiferroic heterostructures\",\"authors\":\"Yiming Sun, Xing Chen, Chao Chen, Baojia Liu, Bingyu Chen, Zhiyuan Zhao, Dahai Wei, Christian H. Back, Wang Kang, Weisheng Zhao, Na Lei\",\"doi\":\"10.1063/5.0221747\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Physical reservoir computing (PRC), a brain-inspired computing method known for its efficient information processing and low training requirements, has attracted significant attention. The key factor lies in the number of computational nodes within the reservoir for its computational capability. Here, we explore co-multiplexing reservoirs that leverage both temporal and spatial strategies. Temporal multiplexing virtually expands the node count through the use of masking techniques, while spatial multiplexing utilizes multiple physical locations (e.g., Hall bars) to achieve an increase in the number of real nodes. Our experiment employs a strain-mediated reservoir based on multiferroic heterostructures. By applying a single voltage across the PMN-PT substrate (acting as global input) and measuring the output Hall voltages from four Hall bars (real nodes), we achieve significant efficiency gains. This co-multiplexing approach results in a reduction in the normalized root mean square error from 0.5 to 0.23 for a 20-step prediction task of a Mackey–Glass chaotic time series. Furthermore, the single input and four independent outputs lead to a fourfold reduction in energy consumption compared to the strain-mediated PRC with temporal multiplexing solely. This research paves the way for future energy saving PRC implementations utilizing co-multiplexing, promoting a resource-efficient paradigm in reservoir computing.\",\"PeriodicalId\":8094,\"journal\":{\"name\":\"Applied Physics Letters\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.5000,\"publicationDate\":\"2024-09-03\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Applied Physics Letters\",\"FirstCategoryId\":\"101\",\"ListUrlMain\":\"https://doi.org/10.1063/5.0221747\",\"RegionNum\":2,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"PHYSICS, APPLIED\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Applied Physics Letters","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1063/5.0221747","RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PHYSICS, APPLIED","Score":null,"Total":0}
Strain-mediated reservoir computing with temporal and spatial co-multiplexing in multiferroic heterostructures
Physical reservoir computing (PRC), a brain-inspired computing method known for its efficient information processing and low training requirements, has attracted significant attention. The key factor lies in the number of computational nodes within the reservoir for its computational capability. Here, we explore co-multiplexing reservoirs that leverage both temporal and spatial strategies. Temporal multiplexing virtually expands the node count through the use of masking techniques, while spatial multiplexing utilizes multiple physical locations (e.g., Hall bars) to achieve an increase in the number of real nodes. Our experiment employs a strain-mediated reservoir based on multiferroic heterostructures. By applying a single voltage across the PMN-PT substrate (acting as global input) and measuring the output Hall voltages from four Hall bars (real nodes), we achieve significant efficiency gains. This co-multiplexing approach results in a reduction in the normalized root mean square error from 0.5 to 0.23 for a 20-step prediction task of a Mackey–Glass chaotic time series. Furthermore, the single input and four independent outputs lead to a fourfold reduction in energy consumption compared to the strain-mediated PRC with temporal multiplexing solely. This research paves the way for future energy saving PRC implementations utilizing co-multiplexing, promoting a resource-efficient paradigm in reservoir computing.
期刊介绍:
Applied Physics Letters (APL) features concise, up-to-date reports on significant new findings in applied physics. Emphasizing rapid dissemination of key data and new physical insights, APL offers prompt publication of new experimental and theoretical papers reporting applications of physics phenomena to all branches of science, engineering, and modern technology.
In addition to regular articles, the journal also publishes invited Fast Track, Perspectives, and in-depth Editorials which report on cutting-edge areas in applied physics.
APL Perspectives are forward-looking invited letters which highlight recent developments or discoveries. Emphasis is placed on very recent developments, potentially disruptive technologies, open questions and possible solutions. They also include a mini-roadmap detailing where the community should direct efforts in order for the phenomena to be viable for application and the challenges associated with meeting that performance threshold. Perspectives are characterized by personal viewpoints and opinions of recognized experts in the field.
Fast Track articles are invited original research articles that report results that are particularly novel and important or provide a significant advancement in an emerging field. Because of the urgency and scientific importance of the work, the peer review process is accelerated. If, during the review process, it becomes apparent that the paper does not meet the Fast Track criterion, it is returned to a normal track.