Nanoscale neural network using non-linear spin-wave interference

Goos-Hänchen shift of inelastically scattered spin-wave beams and cascade nonlinear magnon processes

We study, using micromagnetic simulations, the inelastic scattering of spin-wave beams on edge-localized spin-wave modes in a thin ferromagnetic film. In the splitting and confluence processes, the new spin-wave beams are generated with frequencies shifted by the edge-mode frequency. We report that inelastically scattered spin-wave beams in both processes not only change their direction of propagation but also undergo lateral shifts along the interface, analogous to the Goos-Hänchen effect known in optics. These shifts of inelastically scattered beams, for a few special cases described in the paper, can be in the range of several wavelengths, which is larger than the Goos-Hänchen shift of elastically reflected beam. Unexpectedly, at selected frequencies, we found a significant increase in the value of the lateral shifts of the scattered spin-wave beams formed in the confluence process. We show that this effect is associated with the cascading nonlinear processes taking place at the edge of the film and involving the primary edge spin wave. Our results make an important contribution to the understanding of the nonlinear nature of spin waves and provide a way to exploit it in signal processing with magnons.

Optical control of spin waves in hybrid magnonic-plasmonic structures

Magnonics, which harnesses the unique properties of spin waves, offers promising advancements in data processing due to its broad frequency range, nonlinear dynamics, and scalability for on-chip integration. Effective information encoding in magnonic systems requires precise spatial and temporal control of spin waves. Here, we demonstrate the rapid optical control of spin-wave transport in hybrid magnonic-plasmonic structures. By using thermoplasmonic heating in yttrium iron garnet films integrated with gold nanodisk arrays, we achieve a suppression of spin-wave signals by 20 dB using single laser pulses lasting just a few hundred nanoseconds. Our results reveal a strong correlation between plasmonic light absorption and spin-wave manipulation, as supported by micromagnetic simulations that emphasize the crucial role of magnonic refraction. This study establishes thermoplasmonics as a powerful tool for controlling spin-wave propagation, bridging the fields of magnonics and plasmonics, and paving the way for the development of multifunctional hybrid magnonic devices.

Iono-Magnonic Reservoir Computing With Chaotic Spin Wave Interference Manipulated by Ion-Gating

Physical reservoirs are a promising approach for realizing high-performance artificial intelligence devices utilizing physical devices. Although nonlinear interfered spin-wave multi-detection exhibits high nonlinearity and the ability to map in high dimensional feature space, it does not have sufficient performance to process time-series data precisely. Herein, development of an iono-magnonic reservoir by combining such interfered spin wave multi-detection and ion-gating involving protonation-induced redox reaction triggered by the application of voltage is reported. This study is the first to report the manipulation of the propagating spin wave property by ion-gating and the application of the same to physical reservoir computing. The subject iono-magnonic reservoir can generate various reservoir states in a single homogenous medium by utilizing a spin wave property modulated by ion-gating. Utilizing the strong nonlinearity resulting from chaos, the reservoir shows good computational performance in completing the Mackey-Glass chaotic time-series prediction task, and the performance is comparable to that exhibited by simulated neural networks.

Handedness manipulation of propagating antiferromagnetic magnons

Antiferromagnetic magnons possess a distinctive feature absent in their ferromagnetic counterparts: the presence of two distinct handedness modes, the right-handed (RH) and left-handed (LH) precession modes. The magnon handedness determines the sign of spin polarization carried by the propagating magnon, which is indispensable for harnessing the diverse functionalities in magnonic devices, such as data encoding, magnon polarization-based logic systems, and quantum applications involving magnons. However, the control of coherently propagating magnon handedness in antiferromagnets has remained elusive. Here we demonstrate the manipulation and electrical readout of propagating magnon handedness in perpendicularly magnetized synthetic antiferromagnets (SAF). We find that the antiferromagnetic magnon handedness can be directly identified by measuring the inverse spin Hall effect (ISHE) voltage, which arises from the spin pumping effect caused by the propagating antiferromagnetic magnons in the SAF structure. The RH and LH modes of the magnon can be distinguishable when the SAF structure is sandwiched by heavy metals with the same sign of spin Hall angle. This work unveils promising avenues for harnessing the unique properties of antiferromagnetic magnons.

High data rate spin-wave transmitter

Spin-wave devices have recently become a strong competitor in computing and information processing owing to their excellent energy efficiency. Researchers have explored magnons, the quanta of spin-waves, as an information carrier and significant progress has occurred in both excitation and computation. However, most transmission designs remain immature in terms of data rate and information complexity as they only utilize simple spin-wave pulses and suffer from signal distortion. In this work, using micromagnetic simulations, we demonstrate a spin-wave transmitter that operates reliably at a data rate of 4 Gbps over significant (multi-micron) distances with error rates as low as 10-14. Spin-wave amplitude is used to encode information. Carrier frequency and data rate are carefully chosen to restrict dispersion spreading, which is the main reason for signal distortion. We show that this device can be integrated into either pure-magnonic circuits or modern electronic networks. Our study reveals the potential for achieving an even higher data rate of 10 Gbps and also offers a comprehensive and logical methodology for performance tuning.

Steerable current-driven emission of spin waves in magnetic vortex pairs

The efficient excitation of spin waves is a key challenge in the realization of magnonic devices. We demonstrate current-driven generation of spin waves in antiferromagnetically coupled magnetic vortices. We use time-resolved x-ray microscopy to directly image the emission of spin waves upon the application of alternating currents flowing directly through the magnetic stack. Micromagnetic simulations allow us to identify the current-driven Oersted field as the main origin of excitation, in contrast to spin-transfer torques. In our case, these internal Oersted fields have an orders of magnitude higher spin-wave excitation efficiency than commonly used stripline antennas. For magnetostrictive materials, we furthermore demonstrate that the direction of magnon propagation can be steered by increasing the excitation amplitude, which modifies the underlying magnetization profile through an additional anisotropy. The demonstrated methods allow for the efficient and tunable excitation of spin waves, marking a substantial advance concerning the design of magnonic devices.

All-magnonic repeater based on bistability

Bistability, a universal phenomenon found in diverse fields such as biology, chemistry, and physics, describes a scenario in which a system has two stable equilibrium states and resets to one of the two states. The ability to switch between these two states is the basis for a wide range of applications, particularly in memory and logic operations. Here, we present a universal approach to achieve bistable switching in magnonics, the field processing data using spin waves. A pronounced bistable window is observed in a 1 μm wide magnonic conduit under an external rf drive. The system is characterized by two magnonic stable states defined as low and high spin-wave amplitude states. The switching between these two states is realized by another propagating spin wave sent into the rf driven region. This magnonic bistable switching is used to design a magnonic repeater, which receives the original decayed and distorted spin wave and regenerates a new spin wave with amplified amplitude and normalized phase. Our magnonic repeater can be installed at the inputs of each magnonic logic gate to overcome the spin-wave amplitude degradation and phase distortion during previous propagation and achieve integrated magnonic circuits or magnonic neuromorphic networks.

Neuromorphic overparameterisation and few-shot learning in multilayer physical neural networks

Physical neuromorphic computing, exploiting the complex dynamics of physical systems, has seen rapid advancements in sophistication and performance. Physical reservoir computing, a subset of neuromorphic computing, faces limitations due to its reliance on single systems. This constrains output dimensionality and dynamic range, limiting performance to a narrow range of tasks. Here, we engineer a suite of nanomagnetic array physical reservoirs and interconnect them in parallel and series to create a multilayer neural network architecture. The output of one reservoir is recorded, scaled and virtually fed as input to the next reservoir. This networked approach increases output dimensionality, internal dynamics and computational performance. We demonstrate that a physical neuromorphic system can achieve an overparameterised state, facilitating meta-learning on small training sets and yielding strong performance across a wide range of tasks. Our approach's efficacy is further demonstrated through few-shot learning, where the system rapidly adapts to new tasks.

Emergent coherent modes in nonlinear magnonic waveguides detected at ultrahigh frequency resolution

Nonlinearity of dynamic systems plays a key role in neuromorphic computing, which is expected to reduce the ever-increasing power consumption of machine learning and artificial intelligence applications. For spin waves (magnons), nonlinearity combined with phase coherence is the basis of phenomena like Bose-Einstein condensation, frequency combs, and pattern recognition in neuromorphic computing. Yet, the broadband electrical detection of these phenomena with high-frequency resolution remains a challenge. Here, we demonstrate the generation and detection of phase-coherent nonlinear magnons in an all-electrical GHz probe station based on coplanar waveguides connected to a vector network analyzer which we operate in a frequency-offset mode. Making use of an unprecedented frequency resolution, we resolve the nonlocal emergence of a fine structure of propagating nonlinear magnons, which sensitively depends on both power and a magnetic field. These magnons are shown to maintain coherency with the microwave source while propagating over macroscopic distances. We propose a multi-band four-magnon scattering scheme that is in agreement with the field-dependent characteristics of coherent nonlocal signals in the nonlinear excitation regime. Our findings are key to enable the seamless integration of nonlinear magnon processes into high-speed microwave electronics and to advance phase-encoded information processing in magnonic neuronal networks.

Reversing the magnetization of 50-nm-wide ferromagnets by ultrashort magnons in thin-film yttrium iron garnet

Spin waves (magnons) can enable neuromorphic computing by which one aims at overcoming limitations inherent to conventional electronics and the von Neumann architecture. Encoding magnon signal by reversing magnetization of a nanomagnetic memory bit is pivotal to realize such novel computing schemes efficiently. A magnonic neural network was recently proposed consisting of differently configured nanomagnets that control nonlinear magnon interference in an underlying yttrium iron garnet (YIG) film [Papp et al., Nat. Commun., 2021, 12, 6422]. In this study, we explore the nonvolatile encoding of magnon signals by switching the magnetization of periodic and aperiodic arrays (gratings) of Ni81Fe19 (Py) nanostripes with widths w between 50 nm and 200 nm. Integrating 50-nm-wide nanostripes with a coplanar waveguide, we excited magnons having a wavelength λ of ≈100 nm. At a small spin-precessional power of 11 nW, these ultrashort magnons switch the magnetization of 50-nm-wide Py nanostripes after they have propagated over 25 μm in YIG in an applied field. We also demonstrate the magnetization reversal of nanostripes patterned in an aperiodic sequence. We thereby show that the magnon-induced reversal happens regardless of the width and periodicity of the nanostripe gratings. Our study enlarges substantially the parameter regime for magnon-induced nanomagnet reversal on YIG and is important for realizing in-memory computing paradigms making use of magnons with ultrashort wavelengths at low power consumption.

Tailoring magnon modes by extending square, kagome, and trigonal spin ice lattices vertically via interlayer coupling of trilayer nanomagnets

In this work high-frequency magnetization dynamics and statics of artificial spin-ice lattices with different geometric nanostructure array configurations are studied where the individual nanostructures are composed of ferromagnetic/non-magnetic/ferromagnetic trilayers with different non-magnetic thicknesses. These thickness variations enable additional control over the magnetic interactions within the spin-ice lattice that directly impacts the resulting magnetization dynamics and the associated magnonic modes. Specifically the geometric arrangements studied are square, kagome and trigonal spin ice configurations, where the individual lithographically patterned nanomagnets (NMs) are trilayers, made up of two magnetic layers ofNi81Fe19of 30 nm and 70 nm thickness respectively, separated by a non-magnetic copper layer of either 2 nm or 40 nm. We show that coupling via the magnetostatic interactions between the ferromagnetic layers of the NMs within square, kagome and trigonal spin-ice lattices offers fine-control over magnetization states and magnetic resonant modes. In particular, the kagome and trigonal lattices allow tuning of an additional mode and the spacing between multiple resonance modes, increasing functionality beyond square lattices. These results demonstrate the ability to move beyond quasi-2D single magnetic layer nanomagnetics via control of the vertical interlayer interactions in spin ice arrays. This additional control enables multi-mode magnonic programmability of the resonance spectra, which has potential for magnetic metamaterials for microwave or information processing applications.

The 2024 magnonics roadmap

Magnonicsis a research field that has gained an increasing interest in both the fundamental and applied sciences in recent years. This field aims to explore and functionalize collective spin excitations in magnetically ordered materials for modern information technologies, sensing applications and advanced computational schemes. Spin waves, also known as magnons, carry spin angular momenta that allow for the transmission, storage and processing of information without moving charges. In integrated circuits, magnons enable on-chip data processing at ultrahigh frequencies without the Joule heating, which currently limits clock frequencies in conventional data processors to a few GHz. Recent developments in the field indicate that functional magnonic building blocks for in-memory computation, neural networks and Ising machines are within reach. At the same time, the miniaturization of magnonic circuits advances continuously as the synergy of materials science, electrical engineering and nanotechnology allows for novel on-chip excitation and detection schemes. Such circuits can already enable magnon wavelengths of 50 nm at microwave frequencies in a 5G frequency band. Research into non-charge-based technologies is urgently needed in view of the rapid growth of machine learning and artificial intelligence applications, which consume substantial energy when implemented on conventional data processing units. In its first part, the 2024 Magnonics Roadmap provides an update on the recent developments and achievements in the field of nano-magnonics while defining its future avenues and challenges. In its second part, the Roadmap addresses the rapidly growing research endeavors on hybrid structures and magnonics-enabled quantum engineering. We anticipate that these directions will continue to attract researchers to the field and, in addition to showcasing intriguing science, will enable unprecedented functionalities that enhance the efficiency of alternative information technologies and computational schemes.

Three-dimensional spin-wave dynamics, localization and interference in a synthetic antiferromagnet

Spin waves are collective perturbations in the orientation of the magnetic moments in magnetically ordered materials. Their rich phenomenology is intrinsically three-dimensional; however, the three-dimensional imaging of spin waves has so far not been possible. Here, we image the three-dimensional dynamics of spin waves excited in a synthetic antiferromagnet, with nanoscale spatial resolution and sub-ns temporal resolution, using time-resolved magnetic laminography. In this way, we map the distribution of the spin-wave modes throughout the volume of the structure, revealing unexpected depth-dependent profiles originating from the interlayer dipolar interaction. We experimentally demonstrate the existence of complex three-dimensional interference patterns and analyze them via micromagnetic modelling. We find that these patterns are generated by the superposition of spin waves with non-uniform amplitude profiles, and that their features can be controlled by tuning the composition and structure of the magnetic system. Our results open unforeseen possibilities for the study and manipulation of complex spin-wave modes within nanostructures and magnonic devices.

Magnon-Assisted Magnetization Reversal of Ni81Fe19 Nanostripes on Y3Fe5O12 with Different Interfaces

Magnetic bit writing by short-wave magnons without conversion to the electrical domain is expected to be a game-changer for in-memory computing architectures. Recently, the reversal of nanomagnets by propagating magnons was demonstrated. However, experiments have not yet explored different wavelengths and the nonlinear excitation regime of magnons required for computational tasks. We report on the magnetization reversal of individual 20 nm thick Ni81Fe19 (Py) nanostripes integrated onto 113 nm thick yttrium iron garnet (YIG). We suppress direct interlayer exchange coupling by an intermediate layer, such as Cu and SiO2. By exciting magnons in YIG with wavelengths λ down to 148 nm we observe the reversal of the integrated ferromagnets in a small external field of 14 mT. Magnons with a small wavelength of λ = 195 nm, i.e., twice the width of the Py nanostripes, induced the reversal at a spin-precessional power of only about 1 nW after propagating over 15 μm in YIG. Such small power value has not been reported so far. Considerations based on dynamic dipolar coupling explain the observed wavelength dependence of the magnon-induced reversal efficiency. For an increased power, the stripes reversed in an external field of only about 1 mT. Our findings are important for the practical implementation of nonvolatile storage of broadband magnon signals in YIG by means of bistable nanomagnets without the need of an appreciable global magnetic field.

Three-dimensional magnetic nanotextures with high-order vorticity in soft magnetic wireframes

Additive nanotechnology enable curvilinear and three-dimensional (3D) magnetic architectures with tunable topology and functionalities surpassing their planar counterparts. Here, we experimentally reveal that 3D soft magnetic wireframe structures resemble compact manifolds and accommodate magnetic textures of high order vorticity determined by the Euler characteristic, χ. We demonstrate that self-standing magnetic tetrapods (homeomorphic to a sphere; χ = + 2) support six surface topological solitons, namely four vortices and two antivortices, with a total vorticity of + 2 equal to its Euler characteristic. Alternatively, wireframe structures with one loop (homeomorphic to a torus; χ = 0) possess equal number of vortices and antivortices, which is relevant for spin-wave splitters and 3D magnonics. Subsequent introduction of n holes into the wireframe geometry (homeomorphic to an n-torus; χ < 0) enables the accommodation of a virtually unlimited number of antivortices, which suggests their usefulness for non-conventional (e.g., reservoir) computation. Furthermore, complex stray-field topologies around these objects are of interest for superconducting electronics, particle trapping and biomedical applications.

Resonant generation of propagating second-harmonic spin waves in nano-waveguides

Generation of second-harmonic waves is one of the universal nonlinear phenomena that have found numerous technical applications in many modern technologies, in particular, in photonics. This phenomenon also has great potential in the field of magnonics, which considers the use of spin waves in magnetic nanostructures to implement wave-based signal processing and computing. However, due to the strong frequency dependence of the phase velocity of spin waves, resonant phase-matched generation of second-harmonic spin waves has not yet been achieved in practice. Here, we show experimentally that such a process can be realized using a combination of different modes of nano-sized spin-wave waveguides based on low-damping magnetic insulators. We demonstrate that our approach enables efficient spatially-extended energy transfer between interacting waves, which can be controlled by the intensity of the initial wave and the static magnetic field.

True amplification of spin waves in magnonic nano-waveguides

Magnonic nano-devices exploit magnons - quanta of spin waves - to transmit and process information within a single integrated platform that has the potential to outperform traditional semiconductor-based electronics. The main missing cornerstone of this information nanotechnology is an efficient scheme for the amplification of propagating spin waves. The recent discovery of spin-orbit torque provided an elegant mechanism for propagation losses compensation. While partial compensation of the spin-wave losses has been achieved, true amplification - the exponential increase in the spin-wave intensity during propagation - has so far remained elusive. Here we evidence the operating conditions to achieve unambiguous amplification using clocked nanoseconds-long spin-orbit torque pulses in magnonic nano-waveguides, where the effective magnetization has been engineered to be close to zero to suppress the detrimental magnon scattering. We achieve an exponential increase in the intensity of propagating spin waves up to 500% at a propagation distance of several micrometers.

Task-adaptive physical reservoir computing

Reservoir computing is a neuromorphic architecture that may offer viable solutions to the growing energy costs of machine learning. In software-based machine learning, computing performance can be readily reconfigured to suit different computational tasks by tuning hyperparameters. This critical functionality is missing in 'physical' reservoir computing schemes that exploit nonlinear and history-dependent responses of physical systems for data processing. Here we overcome this issue with a 'task-adaptive' approach to physical reservoir computing. By leveraging a thermodynamical phase space to reconfigure key reservoir properties, we optimize computational performance across a diverse task set. We use the spin-wave spectra of the chiral magnet Cu2OSeO3 that hosts skyrmion, conical and helical magnetic phases, providing on-demand access to different computational reservoir responses. The task-adaptive approach is applicable to a wide variety of physical systems, which we show in other chiral magnets via above (and near) room-temperature demonstrations in Co8.5Zn8.5Mn3 (and FeGe).

On-chip phonon-magnon reservoir for neuromorphic computing

Reservoir computing is a concept involving mapping signals onto a high-dimensional phase space of a dynamical system called "reservoir" for subsequent recognition by an artificial neural network. We implement this concept in a nanodevice consisting of a sandwich of a semiconductor phonon waveguide and a patterned ferromagnetic layer. A pulsed write-laser encodes input signals into propagating phonon wavepackets, interacting with ferromagnetic magnons. The second laser reads the output signal reflecting a phase-sensitive mix of phonon and magnon modes, whose content is highly sensitive to the write- and read-laser positions. The reservoir efficiently separates the visual shapes drawn by the write-laser beam on the nanodevice surface in an area with a size comparable to a single pixel of a modern digital camera. Our finding suggests the phonon-magnon interaction as a promising hardware basis for realizing on-chip reservoir computing in future neuromorphic architectures.

Optically Addressing Exciton Spin and Pseudospin in Nanomaterials for Spintronics Applications

Oriented exciton spins that can be generated and manipulated optically are of interest for a range of applications, including spintronics, quantum information science, and neuromorphic computing architectures. Although materials that host such excitons often lack practical coherence times for use on their own, strategic transduction of the magnetic information across interfaces can combine fast modulation with longer-term storage and readout. Several nanostructure systems have been put forward due to their interesting magneto-optical properties and their possible manipulation using circularly polarized light. These material systems are presented here, namely two-dimensional (2D) systems due to the unique spin-valley coupling properties and quantum dots for their exciton fine structure. 2D magnets are also discussed for their anisotropic spin behavior and extensive 2D magnetic states that are not yet fully understood but could pave the way for emergent techniques of magnetic control. This review also details the experimental and theoretical tools to measure and understand these systems along with a discussion on the progress of optical manipulation of spins and magnetic order transitions.

Zero-Field Spin Waves in YIG Nanowaveguides

Spin-wave-based transmission and processing of information is a promising emerging nanotechnology that can help overcome limitations of traditional electronics based on the transfer of electrical charge. Among the most important challenges for this technology is the implementation of spin-wave devices that can operate without the need for an external bias magnetic field. Here we experimentally demonstrate that this can be achieved using submicrometer wide spin-wave waveguides fabricated from ultrathin films of a low-loss magnetic insulator, yttrium iron garnet (YIG). We show that these waveguides exhibit a highly stable single-domain static magnetic configuration at zero field and support long-range propagation of spin waves with gigahertz frequencies. The experimental results are supported by micromagnetic simulations, which additionally provide information for the optimization of zero-field guiding structures. Our findings create the basis for the development of energy-efficient zero-field spin-wave devices and circuits.

Realization and Control of Bulk and Surface Modes in 3D Nanomagnonic Networks by Additive Manufacturing of Ferromagnets

The high-density integration in information technology fuels the research on functional 3D nanodevices. Particularly ferromagnets promise multifunctional 3D devices for nonvolatile data storage, high-speed data processing, and non-charge-based logic operations via spintronics and magnonics concepts. However, 3D nanofabrication of ferromagnets is extremely challenging. In this work, an additive manufacturing methodology is reported, and unprecedented 3D ferromagnetic nanonetworks with a woodpile-structure unit cell are fabricated. The collective spin dynamics (magnons) at frequencies up to 25 GHz are investigated by Brillouin Light Scattering (BLS) microscopy and micromagnetic simulations. A clear discrepancy of about 10 GHz is found between the bulk and surface modes, which are engineered by different unit cell sizes in the Ni-based nanonetworks. The angle- and spatially-dependent modes demonstrate opportunities for multi-frequency signal processing in 3D circuits via magnons. The developed synthesis route will allow one to create 3D magnonic crystals with chiral unit cells, which are a prerequisite toward surface modes with topologically protected properties.

Deeply nonlinear excitation of self-normalized short spin waves

Spin waves are ideal candidates for wave-based computing, but the construction of magnetic circuits is blocked by a lack of an efficient mechanism to excite long-running exchange spin waves with normalized amplitudes. Here, we solve the challenge by exploiting a deeply nonlinear phenomenon for forward volume spin waves in 200-nm-wide nanoscale waveguides and validate our concept using microfocused Brillouin light scattering spectroscopy. An unprecedented nonlinear frequency shift of more than 2 GHz is achieved, corresponding to a magnetization precession angle of 55° and enabling the excitation of spin waves with wavelengths down to 200 nm. The amplitude of the excited spin waves is constant and independent of the input microwave power due to the self-locking nonlinear shift, enabling robust adjustment of the spin-wave amplitudes in future on-chip magnonic integrated circuits.

magnum.np: a PyTorch based GPU enhanced finite difference micromagnetic simulation framework for high level development and inverse design

magnum.np is a micromagnetic finite-difference library completely based on the tensor library PyTorch. The use of such a high level library leads to a highly maintainable and extensible code base which is the ideal candidate for the investigation of novel algorithms and modeling approaches. On the other hand magnum.np benefits from the device abstraction and optimizations of PyTorch enabling the efficient execution of micromagnetic simulations on a number of computational platforms including graphics processing units and potentially Tensor processing unit systems. We demonstrate a competitive performance to state-of-the-art micromagnetic codes such as mumax3 and show how our code enables the rapid implementation of new functionality. Furthermore, handling inverse problems becomes possible by using PyTorch's autograd feature.

Pattern recognition in reciprocal space with a magnon-scattering reservoir

Magnons are elementary excitations in magnetic materials and undergo nonlinear multimode scattering processes at large input powers. In experiments and simulations, we show that the interaction between magnon modes of a confined magnetic vortex can be harnessed for pattern recognition. We study the magnetic response to signals comprising sine wave pulses with frequencies corresponding to radial mode excitations. Three-magnon scattering results in the excitation of different azimuthal modes, whose amplitudes depend strongly on the input sequences. We show that recognition rates as high as 99.4% can be attained for four-symbol sequences using the scattered modes, with strong performance maintained with the presence of amplitude noise in the inputs.

3D Magnonic Conduits by Direct Write Nanofabrication

Magnonics is a rapidly developing domain of nanomagnetism, with application potential in information processing systems. Realisation of this potential and miniaturisation of magnonic circuits requires their extension into the third dimension. However, so far, magnonic conduits are largely limited to thin films and 2D structures. Here, we introduce 3D magnonic nanoconduits fabricated by the direct write technique of focused-electron-beam induced deposition (FEBID). We use Brillouin light scattering (BLS) spectroscopy to demonstrate significant qualitative differences in spatially resolved spin-wave resonances of 2D and 3D nanostructures, which originates from the geometrically induced non-uniformity of the internal magnetic field. This work demonstrates the capability of FEBID as an additive manufacturing technique to produce magnetic 3D nanoarchitectures and presents the first report of BLS spectroscopy characterisation of FEBID conduits.

Nonlinear multi-magnon scattering in artificial spin ice

Magnons, the quantum-mechanical fundamental excitations of magnetic solids, are bosons whose number does not need to be conserved in scattering processes. Microwave-induced parametric magnon processes, often called Suhl instabilities, have been believed to occur in magnetic thin films only, where quasi-continuous magnon bands exist. Here, we reveal the existence of such nonlinear magnon-magnon scattering processes and their coherence in ensembles of magnetic nanostructures known as artificial spin ice. We find that these systems exhibit effective scattering processes akin to those observed in continuous magnetic thin films. We utilize a combined microwave and microfocused Brillouin light scattering measurement approach to investigate the evolution of their modes. Scattering events occur between resonance frequencies that are determined by each nanomagnet's mode volume and profile. Comparison with numerical simulations reveals that frequency doubling is enabled by exciting a subset of nanomagnets that, in turn, act as nanosized antennas, an effect that is akin to scattering in continuous films. Moreover, our results suggest that tunable directional scattering is possible in these structures.

Spin-Wave Optics in YIG Realized by Ion-Beam Irradiation

Direct focused-ion-beam writing is presented as an enabling technology for realizing functional spin-wave devices of high complexity, and demonstrate its potential by optically-inspired designs. It is shown that ion-beam irradiation changes the characteristics of yttrium iron garnet films on a submicron scale in a highly controlled way, allowing one to engineer the magnonic index of refraction adapted to desired applications. This technique does not physically remove material, and allows rapid fabrication of high-quality architectures of modified magnetization in magnonic media with minimal edge damage (compared to more common removal techniques such as etching or milling). By experimentally showing magnonic versions of a number of optical devices (lenses, gratings, Fourier-domain processors) this technology is envisioned as the gateway to building magnonic computing devices that rival their optical counterparts in their complexity and computational power.

Reversal of nanomagnets by propagating magnons in ferrimagnetic yttrium iron garnet enabling nonvolatile magnon memory

Despite the unprecedented downscaling of CMOS integrated circuits, memory-intensive machine learning and artificial intelligence applications are limited by data conversion between memory and processor. There is a challenging quest for novel approaches to overcome this so-called von Neumann bottleneck. Magnons are the quanta of spin waves. Their angular momentum enables power-efficient computation without charge flow. The conversion problem would be solved if spin wave amplitudes could be stored directly in a magnetic memory. Here, we report the reversal of ferromagnetic nanostripes by spin waves which propagate in an underlying spin-wave bus. Thereby, the charge-free angular momentum flow is stored after transmission over a macroscopic distance. We show that the spin waves can reverse large arrays of ferromagnetic stripes at a strikingly small power level. Combined with the already existing wave logic, our discovery is path-breaking for the new era of magnonics-based in-memory computation and beyond von Neumann computer architectures.

Low-Loss Nanoscopic Spin-Wave Guiding in Continuous Yttrium Iron Garnet Films

Long-distance transport and control of spin waves through nanochannels is essential for integrated magnonic technology. Current strategies relying on the patterning of single-layer nano-waveguides suffer from a decline of the spin-wave decay length upon downscaling or require large magnetic bias field. Here, we introduce a new waveguiding structure based on low-damping continuous yttrium iron garnet (YIG) films. Rather than patterning the YIG film, we define nanoscopic spin-wave transporting channels within YIG by dipolar coupling to ferromagnetic metal nanostripes. The hybrid material structure offers long-distance transport of spin waves with a decay length of ∼20 μm in 160 nm wide waveguides over a broad frequency range at small bias field. We further evidence that spin waves can be redirected easily by stray-field-induced bends in continuous YIG films. The combination of low-loss spin-wave guiding and straightforward nanofabrication highlights a new approach toward the implementation of magnonic integrated circuits for spin-wave computing.

Giant nonlinear self-phase modulation of large-amplitude spin waves in microscopic YIG waveguides

Nonlinear self-phase modulation is a universal phenomenon responsible, for example, for the formation of propagating dynamic solitons. It has been reported for waves of different physical nature. However its direct experimental observation for spin waves has been challenging. Here we show that exceptionally strong phase modulation can be achieved for spin waves in microscopic waveguides fabricated from nanometer-thick films of magnetic insulator, which support propagation of spin waves with large amplitudes corresponding to angles of magnetization precession exceeding 10°. At these amplitudes, the nonstationary nonlinear dynamic response of the spin system causes an extreme broadening of the spectrum of spin-wave pulses resulting in a strong spatial variation of the spin-wave wavelength and a temporal variation of the spin-wave phase across the pulse. Our findings demonstrate great complexity of nonlinear wave processes in microscopic magnetic structures and importance of their understanding for technical applications of spin waves in integrated devices.

Reconfigurable training and reservoir computing in an artificial spin-vortex ice via spin-wave fingerprinting

Strongly interacting artificial spin systems are moving beyond mimicking naturally occurring materials to emerge as versatile functional platforms, from reconfigurable magnonics to neuromorphic computing. Typically, artificial spin systems comprise nanomagnets with a single magnetization texture: collinear macrospins or chiral vortices. By tuning nanoarray dimensions we have achieved macrospin-vortex bistability and demonstrated a four-state metamaterial spin system, the 'artificial spin-vortex ice' (ASVI). ASVI can host Ising-like macrospins with strong ice-like vertex interactions and weakly coupled vortices with low stray dipolar field. Vortices and macrospins exhibit starkly differing spin-wave spectra with analogue mode amplitude control and mode frequency shifts of Δf = 3.8 GHz. The enhanced bitextural microstate space gives rise to emergent physical memory phenomena, with ratchet-like vortex injection and history-dependent non-linear fading memory when driven through global magnetic field cycles. We employed spin-wave microstate fingerprinting for rapid, scalable readout of vortex and macrospin populations, and leveraged this for spin-wave reservoir computation. ASVI performs non-linear mapping transformations of diverse input and target signals in addition to chaotic time-series forecasting.

Forecasting the outcome of spintronic experiments with Neural Ordinary Differential Equations

Deep learning has an increasing impact to assist research, allowing, for example, the discovery of novel materials. Until now, however, these artificial intelligence techniques have fallen short of discovering the full differential equation of an experimental physical system. Here we show that a dynamical neural network, trained on a minimal amount of data, can predict the behavior of spintronic devices with high accuracy and an extremely efficient simulation time, compared to the micromagnetic simulations that are usually employed to model them. For this purpose, we re-frame the formalism of Neural Ordinary Differential Equations to the constraints of spintronics: few measured outputs, multiple inputs and internal parameters. We demonstrate with Neural Ordinary Differential Equations an acceleration factor over 200 compared to micromagnetic simulations for a complex problem - the simulation of a reservoir computer made of magnetic skyrmions (20 minutes compared to three days). In a second realization, we show that we can predict the noisy response of experimental spintronic nano-oscillators to varying inputs after training Neural Ordinary Differential Equations on five milliseconds of their measured response to a different set of inputs. Neural Ordinary Differential Equations can therefore constitute a disruptive tool for developing spintronic applications in complement to micromagnetic simulations, which are time-consuming and cannot fit experiments when noise or imperfections are present. Our approach can also be generalized to other electronic devices involving dynamics.