Reservoir Computing With Spin Waves Excited in a Garnet Film

Nakane, Ryosho; Tanaka, Gouhei; Hirose, Akira

doi:10.1109/access.2018.2794584

Cited by 186 publications

(148 citation statements)

References 28 publications

Supporting

Mentioning

145

Contrasting

Order By: Relevance

“…Spintronic based architectures [10][11][12] are promising candidates for practical RC applications due to their low power usage, strong nonlinearity arising from magnetisation dynamics, and their ability to be scaled down to small sizes. In particular, many studies have already been conducted on spin-torque nano-oscillators with experimental results [10,11,[13][14][15][16] demonstrating promising performance as RC implementations.…”

Section: Introductionmentioning

confidence: 99%

Reservoir Computing Using a Spin-Wave Delay-Line Active-Ring Resonator Based on Yttrium-Iron-Garnet Film

Watt

Kostylev

2020

Phys. Rev. Applied

View full text Add to dashboard Cite

We demonstrate the use of a propagating spin waves for implementing a reservoir computing architecture. Our concept utilises an active ring resonator comprising a magnetic thin film delay line with an integrated feedback loop. These systems exhibit strong nonlinearity and delayed response, two important properties required for an effective reservoir computing implementation. In a simple design, we exploit the electric control of feedback gain to inject input data into the active ring resonator and use a microwave diode to read out the amplitude of the spin waves circulating in the ring. We employ two baseline tasks, namely the short term memory and parity check tasks, to evaluate the suitability of this architecture for processing time series data.

show abstract

Section: Introductionmentioning

confidence: 99%

Reservoir Computing Using a Spin-Wave Delay-Line Active-Ring Resonator Based on Yttrium-Iron-Garnet Film

Watt

Kostylev

2020

Phys. Rev. Applied

View full text Add to dashboard Cite

show abstract

“…In this context, surface acoustic waves and bulk acoustic waves have represented reference technologies for wave‐based wireless communication, but their scalability is limited . On the other hand, spin waves offer the potential for nanoscale integrability and provide an interesting physical system for developing unconventional computational frameworks, such as neural networks and reservoir computing …”

mentioning

confidence: 99%

“…[6] On the other hand, spin waves offer the potential for nanoscale integrability and provide an interesting physical system for developing unconventional computational frameworks, such as neural networks and reservoir computing. [7] To exploit the rich phenomenology of spin waves [8][9][10][11][12][13][14] for integrated optically inspired processing, generating coherent spatially engineered wavefronts and controlling the propagation and interference of multiple spin-wave beams are crucial. In addition, nonreciprocity, arising from the dipolar interactions, [15][16][17][18] nonreciprocal coupling between spin waves and antennas, [12] and the breaking of the top/ bottom symmetry of the ferromagnetic films, [19] represents an additional degree of freedom for the realization of devices.One of the most versatile methods for spin-wave emission is based on using patterned shaped microantennas, for generating a localized oscillating magnetic field in correspondence Integrated optically inspired wave-based processing is envisioned to outperform digital architectures in specific tasks, such as image processing and speech recognition.…”

mentioning

confidence: 99%

Optically Inspired Nanomagnonics with Nonreciprocal Spin Waves in Synthetic Antiferromagnets

et al. 2020

View full text Add to dashboard Cite

Integrated optically inspired wave‐based processing is envisioned to outperform digital architectures in specific tasks, such as image processing and speech recognition. In this view, spin waves represent a promising route due to their nanoscale wavelength in the gigahertz frequency range and rich phenomenology. Here, a versatile, optically inspired platform using spin waves is realized, demonstrating the wavefront engineering, focusing, and robust interference of spin waves with nanoscale wavelength. In particular, magnonic nanoantennas based on tailored spin textures are used for launching spatially shaped coherent wavefronts, diffraction‐limited spin‐wave beams, and generating robust multi‐beam interference patterns, which spatially extend for several times the spin‐wave wavelength. Furthermore, it is shown that intriguing features, such as resilience to back reflection, naturally arise from the spin‐wave nonreciprocity in synthetic antiferromagnets, preserving the high quality of the interference patterns from spurious counterpropagating modes. This work represents a fundamental step toward the realization of nanoscale optically inspired devices based on spin waves.

show abstract

“…Manipulating the spin degree of freedom in solid-state matter is a promising route for brain-inspired computing [8,9], due to the combination of (i) high-quality materials available to which individual and coupled moments can be manipulated down to the atomic scale, (ii) rich landscape of non-linear, dynamic, and stochastic spin-based phenomena, (iii) the variety of read/write options available. Recently, many schemes have been proposed which utilize the spin degree of freedom in hardware, to perform machine learning tasks [10][11][12][13]. Of particular interest is the Hopfield network: a recurrent neural network that implements an associative memory [14].…”

Section: Introductionmentioning

confidence: 99%

Atom-by-atom construction of attractors in a tunable finite size spin array

et al. 2020

View full text Add to dashboard Cite

We demonstrate that a two-dimensional finite and periodic array of Ising spins coupled via RKKY-like exchange can exhibit tunable magnetic states ranging across three distinct magnetic regimes: (1) a conventional ferromagnetic regime, (2) a glass-like regime, and (3) a new multi-well regime. These magnetic regimes can be tuned by one gate-like parameter, namely the ratio between the lattice constant and the oscillating interaction wavelength. We characterize the various magnetic regimes, quantifying the distribution of low energy states, aging relaxation dynamics, and scaling behavior. The glassy and multi-well behavior results from the competing character of the oscillating long-range exchange interactions with respect to the lattice. The multi-well structure features multiple attractors, each with a sizable basin of attraction. This may open the possible application of such atomic arrays as associative memories.

show abstract

Reservoir Computing With Spin Waves Excited in a Garnet Film

Cited by 186 publications

References 28 publications

Reservoir Computing Using a Spin-Wave Delay-Line Active-Ring Resonator Based on Yttrium-Iron-Garnet Film

Reservoir Computing Using a Spin-Wave Delay-Line Active-Ring Resonator Based on Yttrium-Iron-Garnet Film

Optically Inspired Nanomagnonics with Nonreciprocal Spin Waves in Synthetic Antiferromagnets

Atom-by-atom construction of attractors in a tunable finite size spin array

Contact Info

Product

Resources

About