Pub Date : 2024-07-24DOI: 10.1088/2634-4386/ad6732
Wei Wang, Yang Li, Minghua Wang
Crossbar arrays of memristors are promising to accelerate the deep learning algorithm as a non-von-Neumann architecture, where the computation happens at the location of the memory. The computations are parallelly conducted employing the basic physical laws. However, current research works mainly focus on the offline training of deep neural networks, i.e., only the information forwarding is accelerated by the crossbar arrays. Two other essential operations, i.e., error backpropagation and weight update, are mostly simulated and coordinated by a conventional computer in von Neumann architecture, respectively. Several different in situ learning schemes incorporating error backpropagation and/or weight updates have been proposed and investigated through simulation. Nevertheless, they met the issues of non-ideal synaptic behaviors of the memristors and the complexities of the neural circuits surrounding crossbar arrays. Here we review the difficulties in implementing the error backpropagation and weight update operations for online training or in-memory learning that are adapted to noisy and non-ideal memristors. We hope this work will bridge the gap between the device engineers who are struggling to develop an ideal synaptic device and neural network algorithmists who are assuming that ideal devices are right at hand. The close of this gap could push forward the information processing system paradigm from computing-in-memory to learning-in-memory, aiming at a standalone non-von-Neumann computing system.
{"title":"Difficulties and approaches in enabling learning-in-memory using crossbar arrays of memristors","authors":"Wei Wang, Yang Li, Minghua Wang","doi":"10.1088/2634-4386/ad6732","DOIUrl":"https://doi.org/10.1088/2634-4386/ad6732","url":null,"abstract":"\u0000 Crossbar arrays of memristors are promising to accelerate the deep learning algorithm as a non-von-Neumann architecture, where the computation happens at the location of the memory. The computations are parallelly conducted employing the basic physical laws. However, current research works mainly focus on the offline training of deep neural networks, i.e., only the information forwarding is accelerated by the crossbar arrays. Two other essential operations, i.e., error backpropagation and weight update, are mostly simulated and coordinated by a conventional computer in von Neumann architecture, respectively. Several different in situ learning schemes incorporating error backpropagation and/or weight updates have been proposed and investigated through simulation. Nevertheless, they met the issues of non-ideal synaptic behaviors of the memristors and the complexities of the neural circuits surrounding crossbar arrays. Here we review the difficulties in implementing the error backpropagation and weight update operations for online training or in-memory learning that are adapted to noisy and non-ideal memristors. We hope this work will bridge the gap between the device engineers who are struggling to develop an ideal synaptic device and neural network algorithmists who are assuming that ideal devices are right at hand. The close of this gap could push forward the information processing system paradigm from computing-in-memory to learning-in-memory, aiming at a standalone non-von-Neumann computing system.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"65 19","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141806412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-18DOI: 10.1088/2634-4386/ad5fb2
Dingchen Wang, Anran Yuan, Shilei Dai, Xiao Tang, Kunbin Huang, Songrui Wei, Han Zhang, Zhongrui Wang
In the era of the Internet of Things, photonic neuromorphic computing presents a promising method for real-time, local processing of vast quantities of data. However, the rigidity of materials used in such devices can considerably impact performance and longevity when subjected to mechanical deformation. In this study, we introduce a liquid optical memristor (LOM) based on an organic-inorganic hybrid in a liquid state. This novel approach offers programmable optical properties and significant mechanical flexibility thanks to the robust photochromic and capillary effects. We have developed a LOM with a 24 dB cm−1 modulation depth and over 3-bit nonvolatile memory states. By controlling the droplet morphology to mimic a synapse-like shape, the LOM can withstand strains over 400% and endure misalignment and bending. Furthermore, our findings substantiate the application of LOM for photonic neuromorphic computing systems, yielding 100% accuracy in pattern recognition. The easily-integratable LOM paves the way for the creation of flexible and wearable photonic neuromorphic computing systems.
{"title":"A liquid optical memristor using photochromic effect and capillary effect","authors":"Dingchen Wang, Anran Yuan, Shilei Dai, Xiao Tang, Kunbin Huang, Songrui Wei, Han Zhang, Zhongrui Wang","doi":"10.1088/2634-4386/ad5fb2","DOIUrl":"https://doi.org/10.1088/2634-4386/ad5fb2","url":null,"abstract":"\u0000 In the era of the Internet of Things, photonic neuromorphic computing presents a promising method for real-time, local processing of vast quantities of data. However, the rigidity of materials used in such devices can considerably impact performance and longevity when subjected to mechanical deformation. In this study, we introduce a liquid optical memristor (LOM) based on an organic-inorganic hybrid in a liquid state. This novel approach offers programmable optical properties and significant mechanical flexibility thanks to the robust photochromic and capillary effects. We have developed a LOM with a 24 dB cm−1 modulation depth and over 3-bit nonvolatile memory states. By controlling the droplet morphology to mimic a synapse-like shape, the LOM can withstand strains over 400% and endure misalignment and bending. Furthermore, our findings substantiate the application of LOM for photonic neuromorphic computing systems, yielding 100% accuracy in pattern recognition. The easily-integratable LOM paves the way for the creation of flexible and wearable photonic neuromorphic computing systems.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":" 48","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141827346","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-16DOI: 10.1088/2634-4386/ad63c6
Daniela Rana, Chihyeong Kim, Meijing Wang, Fabio Cicoira, F. Santoro
Organic neuromorphic devices are rapidly developing as platforms for computing, automation and biointerfacing. Resembling short- and long-term synaptic plasticity is a key characteristic to create functional neuromorphic interfaces showcasing spiking activity and learning capabilities. This further enables these devices for coupling with biological systems, such as living neuronal cells and ultimately the brain. However, this would require electrochemical neuromorphic organic devices (ENODes) to interface gel-like electrolytes where neurotransmitter can freely diffuse. To this end, we investigated how planar ENODes (electrochemical transistors) with different geometries and based on different PEDOT:PSS formulations can feature short-and long-term plasticity when in contact with diverse tissue-like gel electrolytes containing catecholamine neurotransmitters. We find both the composition of the bulk electrolyte and gate material play a crucial role in diffusion and trapping of cations that ultimately modulate the conductance of the transistor channels. Our work on ENODe-gel coupling could pave the way to effective brain interfacing for computing and neuroelectronic applications.
{"title":"Tissue-like interfacing of planar electrochemical organic neuromorphic devices","authors":"Daniela Rana, Chihyeong Kim, Meijing Wang, Fabio Cicoira, F. Santoro","doi":"10.1088/2634-4386/ad63c6","DOIUrl":"https://doi.org/10.1088/2634-4386/ad63c6","url":null,"abstract":"\u0000 Organic neuromorphic devices are rapidly developing as platforms for computing, automation and biointerfacing. Resembling short- and long-term synaptic plasticity is a key characteristic to create functional neuromorphic interfaces showcasing spiking activity and learning capabilities. This further enables these devices for coupling with biological systems, such as living neuronal cells and ultimately the brain. However, this would require electrochemical neuromorphic organic devices (ENODes) to interface gel-like electrolytes where neurotransmitter can freely diffuse. To this end, we investigated how planar ENODes (electrochemical transistors) with different geometries and based on different PEDOT:PSS formulations can feature short-and long-term plasticity when in contact with diverse tissue-like gel electrolytes containing catecholamine neurotransmitters. We find both the composition of the bulk electrolyte and gate material play a crucial role in diffusion and trapping of cations that ultimately modulate the conductance of the transistor channels. Our work on ENODe-gel coupling could pave the way to effective brain interfacing for computing and neuroelectronic applications.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"50 12","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141643722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-05DOI: 10.1088/2634-4386/ad3a94
D. Woo, Gichang Noh, E. Park, Min Jee Kim, Dae Kyu Lee, Yong Woo Sung, Jaewook Kim, Yeonjoo Jeong, Jongkil Park, Seong Gon Park, Hyun Jae Jang, Nakwon Choi, Y. Jo, J. Y. Kwak
In-memory computing facilitates efficient parallel computing based on the programmable memristor crossbar array. Proficient hardware image processing can be implemented by utilizing the analog vector-matrix operation with multiple memory states of the nonvolatile memristor in the crossbar array. Among various materials, 2D materials are great candidates for a switching layer of nonvolatile memristors, demonstrating low-power operation and electrical tunability through their remarkable physical and electrical properties. However, the intrinsic device-to-device (D2D) variation of memristors within the crossbar array can degrade the accuracy and performance of in-memory computing. Here, we demonstrate hardware image processing using the fabricated 2D hexagonal boron nitride-based memristor to investigate the effects of D2D variation on the hardware convolution process. The image quality is evaluated by peak-signal-to-noise ratio, structural similarity index measure, and Pratt’s figure of merit and analyzed according to D2D variations. Then, we propose a novel two-step gradual reset programming scheme to enhance the conductance uniformity of multiple states of devices. This approach can enhance the D2D variation and demonstrate the improved quality of the image processing result. We believe that this result suggests the precise tuning method to realize high-performance in-memory computing.
{"title":"Implementation of two-step gradual reset scheme for enhancing state uniformity of 2D hBN-based memristors for image processing","authors":"D. Woo, Gichang Noh, E. Park, Min Jee Kim, Dae Kyu Lee, Yong Woo Sung, Jaewook Kim, Yeonjoo Jeong, Jongkil Park, Seong Gon Park, Hyun Jae Jang, Nakwon Choi, Y. Jo, J. Y. Kwak","doi":"10.1088/2634-4386/ad3a94","DOIUrl":"https://doi.org/10.1088/2634-4386/ad3a94","url":null,"abstract":"In-memory computing facilitates efficient parallel computing based on the programmable memristor crossbar array. Proficient hardware image processing can be implemented by utilizing the analog vector-matrix operation with multiple memory states of the nonvolatile memristor in the crossbar array. Among various materials, 2D materials are great candidates for a switching layer of nonvolatile memristors, demonstrating low-power operation and electrical tunability through their remarkable physical and electrical properties. However, the intrinsic device-to-device (D2D) variation of memristors within the crossbar array can degrade the accuracy and performance of in-memory computing. Here, we demonstrate hardware image processing using the fabricated 2D hexagonal boron nitride-based memristor to investigate the effects of D2D variation on the hardware convolution process. The image quality is evaluated by peak-signal-to-noise ratio, structural similarity index measure, and Pratt’s figure of merit and analyzed according to D2D variations. Then, we propose a novel two-step gradual reset programming scheme to enhance the conductance uniformity of multiple states of devices. This approach can enhance the D2D variation and demonstrate the improved quality of the image processing result. We believe that this result suggests the precise tuning method to realize high-performance in-memory computing.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"114 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141674209","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-03DOI: 10.1088/2634-4386/ad5eb5
Ui-Chan Jeong, Jun-Seok Ro, Hea-Lim Park, Tae-Woo Lee
Neuromorphic devices that emulate biological neural systems have been actively studied to overcome the limitations of conventional von Neumann computing structure. Implementing various synaptic characteristics and decay time in the devices is important for various wearable neuromorphic applications. Polymer-based artificial synapses have been proposed as a solution to satisfy these requirements. Owing to the characteristics of polymer conjugated materials, such as easily tunable optical/electrical properties, mechanical flexibility, and biocompatibility, polymer-based synaptic devices are investigated to demonstrate their ultimate applications replicating biological nervous systems. In this review, we discuss various synaptic properties of artificial synaptic devices, including the operating mechanisms of synaptic devices. Furthermore, we review recent studies on polymer-based synaptic devices, focusing on strategies that modulate synaptic plasticity and synaptic decay time by changing the polymer structure and fabrication process. Finally, we show how the modulation of the synaptic properties can be applied to three major categories of these devices, including neuromorphic computing, artificial synaptic devices with sensing functions, and artificial nerves for neuroprostheses.
{"title":"Modulating short-term and long-term plasticity of polymer-based artificial synapses for neuromorphic computing and beyond","authors":"Ui-Chan Jeong, Jun-Seok Ro, Hea-Lim Park, Tae-Woo Lee","doi":"10.1088/2634-4386/ad5eb5","DOIUrl":"https://doi.org/10.1088/2634-4386/ad5eb5","url":null,"abstract":"\u0000 Neuromorphic devices that emulate biological neural systems have been actively studied to overcome the limitations of conventional von Neumann computing structure. Implementing various synaptic characteristics and decay time in the devices is important for various wearable neuromorphic applications. Polymer-based artificial synapses have been proposed as a solution to satisfy these requirements. Owing to the characteristics of polymer conjugated materials, such as easily tunable optical/electrical properties, mechanical flexibility, and biocompatibility, polymer-based synaptic devices are investigated to demonstrate their ultimate applications replicating biological nervous systems. In this review, we discuss various synaptic properties of artificial synaptic devices, including the operating mechanisms of synaptic devices. Furthermore, we review recent studies on polymer-based synaptic devices, focusing on strategies that modulate synaptic plasticity and synaptic decay time by changing the polymer structure and fabrication process. Finally, we show how the modulation of the synaptic properties can be applied to three major categories of these devices, including neuromorphic computing, artificial synaptic devices with sensing functions, and artificial nerves for neuroprostheses.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"67 s306","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141683303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-14DOI: 10.1088/2634-4386/ad5829
Wei Lu, Melika Payvand, Yuch-Chi Yang
Neuromorphic technologies aim to use the organizing principles of the brain to build efficient and intelligent systems, making them the center-piece between the biological and current Artificial Intelligence (AI) systems. Specifically, in conventional AI systems, one of the dominant sources of power consumption is the data movement between the memory and the processor units, known as the von Neumann bottleneck. In-memory computing solves this problem by co-locating memory and processing units, drastically reducing the power as the data are processed where they reside.
{"title":"Editorial: Focus Issue on In-Memory Computing","authors":"Wei Lu, Melika Payvand, Yuch-Chi Yang","doi":"10.1088/2634-4386/ad5829","DOIUrl":"https://doi.org/10.1088/2634-4386/ad5829","url":null,"abstract":"\u0000 Neuromorphic technologies aim to use the organizing principles of the brain to build efficient and intelligent systems, making them the center-piece between the biological and current Artificial Intelligence (AI) systems. Specifically, in conventional AI systems, one of the dominant sources of power consumption is the data movement between the memory and the processor units, known as the von Neumann bottleneck. In-memory computing solves this problem by co-locating memory and processing units, drastically reducing the power as the data are processed where they reside.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"21 9","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141341322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-13DOI: 10.1088/2634-4386/ad57e7
R. W. Ahmad, Rainer Waser, Florian Maudet, Onur Toprak, Catherine Dubourdieu, S. Menzel
Resistively switching electrochemical metallization memory (ECM) cells are gaining huge interest, as they are seen as promising candidates and basic building blocks of future computation-in-memory applications. However, especially filamentary-based memristive devices suffer from inherent variability, originating from their stochastic switching behaviour. A variability-aware compact model of Electrochemical Metallization Memory Cells is presented in this work and verified by showing a fit to experimental data. It is an extension of a deterministic model. As this extension consists of several different features allowing for a realistic variability-aware fit, it depicts a unique model comprising physics-based, stochastically and experimentally originating variabilities and reproduces them well. Also, a physics-based model parameter study is executed, which enables a comprehensive view into the device physics and presents guidelines for the compact model fitting procedure.
{"title":"Variability-aware modelling of electrochemical metallization memory cells","authors":"R. W. Ahmad, Rainer Waser, Florian Maudet, Onur Toprak, Catherine Dubourdieu, S. Menzel","doi":"10.1088/2634-4386/ad57e7","DOIUrl":"https://doi.org/10.1088/2634-4386/ad57e7","url":null,"abstract":"\u0000 Resistively switching electrochemical metallization memory (ECM) cells are gaining huge interest, as they are seen as promising candidates and basic building blocks of future computation-in-memory applications. However, especially filamentary-based memristive devices suffer from inherent variability, originating from their stochastic switching behaviour. A variability-aware compact model of Electrochemical Metallization Memory Cells is presented in this work and verified by showing a fit to experimental data. It is an extension of a deterministic model. As this extension consists of several different features allowing for a realistic variability-aware fit, it depicts a unique model comprising physics-based, stochastically and experimentally originating variabilities and reproduces them well. Also, a physics-based model parameter study is executed, which enables a comprehensive view into the device physics and presents guidelines for the compact model fitting procedure.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141345511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-06DOI: 10.1088/2634-4386/ad54ea
Jiahe Hu, Shangda Qu, Honghuan Xu, Lin Sun, C. Jiang, Lu Yang, Yi Du, Wentao Xu
Mimicking the functional traits of the muscle system evolves the development of the neuromorphic prosthetic limbs. Herein, a bioinspired neuromuscular system was constructed by connecting an information processor with the aid of a flexible electro-optical synaptic transistor (FNST) to an effector that uses artificial muscle fibers. In this system, the response of artificial muscle fibers, which imitates the movement of biological muscle fibers, is manipulated by neuromorphic synaptic devices. The FNST is regulated by light pulses and electrical spikes to emulate biological synaptic functions, and thereby applied in secure communication. The feasibility of n-type organic nanowires acting as the channels for neuromorphic devices was demonstrated. Attributing to the flexibility of the n-type organic semiconductor N2200 nanowires, the current of the FNST retains > 85% of its initial value after the 5000 bending cycles to radius = 1 cm. The tolerance of bending of the FNST implies its potential applications in wearable electronics. This work offers an approach to potentially advancing electronic skin, neuro-controlled robots, and neuromorphic prosthetic limbs.
{"title":"A bioinspired neuromuscular system enabled by flexible electro-optical N2200 nanowire synaptic transistor","authors":"Jiahe Hu, Shangda Qu, Honghuan Xu, Lin Sun, C. Jiang, Lu Yang, Yi Du, Wentao Xu","doi":"10.1088/2634-4386/ad54ea","DOIUrl":"https://doi.org/10.1088/2634-4386/ad54ea","url":null,"abstract":"\u0000 Mimicking the functional traits of the muscle system evolves the development of the neuromorphic prosthetic limbs. Herein, a bioinspired neuromuscular system was constructed by connecting an information processor with the aid of a flexible electro-optical synaptic transistor (FNST) to an effector that uses artificial muscle fibers. In this system, the response of artificial muscle fibers, which imitates the movement of biological muscle fibers, is manipulated by neuromorphic synaptic devices. The FNST is regulated by light pulses and electrical spikes to emulate biological synaptic functions, and thereby applied in secure communication. The feasibility of n-type organic nanowires acting as the channels for neuromorphic devices was demonstrated. Attributing to the flexibility of the n-type organic semiconductor N2200 nanowires, the current of the FNST retains > 85% of its initial value after the 5000 bending cycles to radius = 1 cm. The tolerance of bending of the FNST implies its potential applications in wearable electronics. This work offers an approach to potentially advancing electronic skin, neuro-controlled robots, and neuromorphic prosthetic limbs.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"4 6","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141380480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-04DOI: 10.1088/2634-4386/ad53f9
G. Venkat, Ian T. Vidamour, C. Swindells, Paul W Fry, Mark Rosamond, Michael Foerster, Miguel Angel Niño, David Griffin, Susan Stepney, D. Allwood, T. Hayward
Physical reservoir computing (RC) is a machine learning technique that is ideal for processing of time dependent data series. It is also uniquely well-aligned to in materio computing realisations that allow the inherent memory and non-linear responses of functional materials to be directly exploited for computation. We have previously shown that square arrays of interconnected magnetic nanorings are attractive candidates for in materio reservoir computing, and experimentally demonstrated their strong performance in a range of benchmark tasks. Here, we extend these studies to other lattice arrangements of rings, including trigonal and Kagome grids, to explore how these affect both the magnetic behaviours of the arrays, and their computational properties. We show that while lattice geometry substantially affects the microstate behaviour of the arrays, these differences manifest less profoundly when averaging magnetic behaviour across the arrays. Consequently the computational properties (as measured using task agnostic metrics) of devices with a single electrical readout are found to be only subtly different, with the approach used to time-multiplex data into and out of the arrays having a stronger effect on properties than the lattice geometry. However, we also find that hybrid reservoirs that combine the outputs from arrays with different lattice geometries show enhanced computational properties compared to any single array.
{"title":"Exploring physical and digital architectures in magnetic nanoring array reservoir computers","authors":"G. Venkat, Ian T. Vidamour, C. Swindells, Paul W Fry, Mark Rosamond, Michael Foerster, Miguel Angel Niño, David Griffin, Susan Stepney, D. Allwood, T. Hayward","doi":"10.1088/2634-4386/ad53f9","DOIUrl":"https://doi.org/10.1088/2634-4386/ad53f9","url":null,"abstract":"\u0000 Physical reservoir computing (RC) is a machine learning technique that is ideal for processing of time dependent data series. It is also uniquely well-aligned to in materio computing realisations that allow the inherent memory and non-linear responses of functional materials to be directly exploited for computation. We have previously shown that square arrays of interconnected magnetic nanorings are attractive candidates for in materio reservoir computing, and experimentally demonstrated their strong performance in a range of benchmark tasks. Here, we extend these studies to other lattice arrangements of rings, including trigonal and Kagome grids, to explore how these affect both the magnetic behaviours of the arrays, and their computational properties. We show that while lattice geometry substantially affects the microstate behaviour of the arrays, these differences manifest less profoundly when averaging magnetic behaviour across the arrays. Consequently the computational properties (as measured using task agnostic metrics) of devices with a single electrical readout are found to be only subtly different, with the approach used to time-multiplex data into and out of the arrays having a stronger effect on properties than the lattice geometry. However, we also find that hybrid reservoirs that combine the outputs from arrays with different lattice geometries show enhanced computational properties compared to any single array.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"13 2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141266935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-14DOI: 10.1088/2634-4386/ad4b5b
M. Hejda, E. A. Doris, S. Bilodeau, J. Robertson, D. Owen-Newns, B. Shastri, Paul R. Prucnal, Antonio Hurtado
Spiking neurons and neural networks constitute a fundamental building block for brain-inspired computing, which is poised to benefit significantly from photonic hardware implementations. In this work, we experimentally investigate an interconnected optical neuromorphic system based on an ultrafast spiking vertical cavity surface emitting laser (VCSEL) neuron and a silicon photonics (SiPh) integrated micro-ring resonator (MRR). We experimentally demonstrate two different functional arrangements of these devices: first, we show that MRR weight banks can be used in conjunction with the spiking VCSEL-neurons to perform amplitude weighting of sub-ns optical spiking signals. Second, we show that a continuously firing VCSEL-neuron can be directly modulated using a locking signal propagated through a single weighting MRR, and we utilize this functionality to perform optical spike firing rate-coding via thermal tuning of the MRR. Given the significant track record of both integrated weight banks and photonic VCSEL-neurons, we believe these results demonstrate the viability of combining these two classes of devices for use in functional neuromorphic photonic systems.
{"title":"Optical spike amplitude weighting and neuromimetic rate coding using a joint VCSEL-MRR neuromorphic photonic system","authors":"M. Hejda, E. A. Doris, S. Bilodeau, J. Robertson, D. Owen-Newns, B. Shastri, Paul R. Prucnal, Antonio Hurtado","doi":"10.1088/2634-4386/ad4b5b","DOIUrl":"https://doi.org/10.1088/2634-4386/ad4b5b","url":null,"abstract":"\u0000 Spiking neurons and neural networks constitute a fundamental building block for brain-inspired computing, which is poised to benefit significantly from photonic hardware implementations. In this work, we experimentally investigate an interconnected optical neuromorphic system based on an ultrafast spiking vertical cavity surface emitting laser (VCSEL) neuron and a silicon photonics (SiPh) integrated micro-ring resonator (MRR). We experimentally demonstrate two different functional arrangements of these devices: first, we show that MRR weight banks can be used in conjunction with the spiking VCSEL-neurons to perform amplitude weighting of sub-ns optical spiking signals. Second, we show that a continuously firing VCSEL-neuron can be directly modulated using a locking signal propagated through a single weighting MRR, and we utilize this functionality to perform optical spike firing rate-coding via thermal tuning of the MRR. Given the significant track record of both integrated weight banks and photonic VCSEL-neurons, we believe these results demonstrate the viability of combining these two classes of devices for use in functional neuromorphic photonic systems.","PeriodicalId":198030,"journal":{"name":"Neuromorphic Computing and Engineering","volume":"26 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140980332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：