arXiv subjects

Holly H. Holder

Publications and source records attributed to Holly H. Holder.

4 recordsLinked to original sources

arXiv2024

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.

cond-mat.mes-hall↗

arXiv2024

Ultrastrong magnon-magnon coupling and chiral spin-texture control in a dipolar 3D multilayered artificial spin-vortex ice

Strongly-interacting nanomagnetic arrays are ideal systems for exploring reconfigurable magnonics. They provide huge microstate spaces and integrated solutions for storage and neuromorphic computing alongside GHz functionality. These systems may be broadly assessed by their range of reliably accessible states and the strength of magnon coupling phenomena and nonlinearities. Increasingly, nanomagnetic systems are expanding into three-dimensional architectures. This has enhanced the range of available magnetic microstates and functional behaviours, but engineering control over 3D states and dynamics remains challenging. Here, we introduce a 3D magnonic metamaterial composed from multilayered artificial spin ice nanoarrays. Comprising two magnetic layers separated by a non-magnetic spacer, each nanoisland may assume four macrospin or vortex states per magnetic layer. This creates a system with a rich $16^N$ microstate space and intense static and dynamic dipolar magnetic coupling. The system exhibits a broad range of emergent phenomena driven by the strong inter-layer dipolar interaction, including ultrastrong magnon-magnon coupling with normalised coupling rates of $\frac{Δω}γ = 0.57$, GHz mode shifts in zero applied field and chirality-selective magneto-toroidal microstate programming and corresponding magnonic spectral control.

cond-mat.mes-hall↗

arXiv2022

Low power continuous-wave all-optical magnetic switching in ferromagnetic nanoarrays

All-optical magnetic switching promises ultrafast, high-resolution magnetisation control with the technological attraction of requiring no magnetic field. Existing all-optical switching schemes are driven by ultrafast transient effects, typically requiring power-hungry femtosecond-pulsed lasers and complex magnetic materials. Here, we demonstrate deterministic, all-optical magnetic switching in simple ferromagnetic nanomagnets (Ni$_{81}$Fe$_{19}$, Ni$_{50}$Fe$_{50}$) with sub-diffraction limit dimensions using a focused low-power, linearly-polarised continuous-wave laser. Isolated nanomagnets are switched across a range of dimensions, laser wavelengths and powers. All square-geometry artificial spin ice vertex configurations are written, including ground-state and energetically-unfavourable `monopole-like' states at powers as low as 2.74 mW. Usually, magnetic switching with linearly polarised light is symmetry-forbidden; however, here the laser spot has a similar size to the nanomagnets, producing an absorption distribution dependent on the relative nanoisland-spot displacement. We attribute the observed deterministic switching to the transient dynamics of this asymmetric absorption. No switching is observed in Co samples, suggesting the multi-species nature of NiFe alloys plays a role in reversal. The results presented here usher in cheap, low-power optically-controlled devices with impact across data storage, neuromorphic computation and reconfigurable magnonics.

cond-mat.mes-hall↗

arXiv2022

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 magnetisation texture: collinear macrospins or chiral vortices. By tuning nanoarray dimensions we achieve macrospin/vortex bistability and demonstrate a four-state metamaterial spin-system '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-style mode-amplitude control and mode-frequency shifts of df = 3.8 GHz. The enhanced bi-textural microstate space gives rise to emergent physical memory phenomena, with ratchet-like vortex training and history-dependent nonlinear fading memory when driven through global field cycles. We employ spin-wave microstate fingerprinting for rapid, scaleable readout of vortex and macrospin populations and leverage this for spin-wave reservoir computation. ASVI performs linear and non-linear mapping transformations of diverse input signals as well as chaotic time-series forecasting. Energy costs of machine learning are spiralling unsustainably, developing low-energy neuromorphic computation hardware such as ASVI is crucial to achieving a zero-carbon computational future.

cond-mat.mes-hall↗