16 September 2024 in Nature Communications
Physical reservoir computing leverages the dynamical properties of complex physical systems to process information efficiently, significantly reducing training efforts and energy consumption. Magnetic skyrmions, topological spin textures, are promising candidates for reservoir computing systems due to their enhanced stability, non-linear interactions and low-power manipulation. Traditional spin-based reservoir computing has been limited to quasi-static detection or real-world data must be rescaled to the intrinsic timescale of the reservoir. We address this challenge by time-multiplexed skyrmion reservoir computing, that allows for aligning the reservoir’s intrinsic timescales to real-world temporal patterns. Using millisecond-scale hand gestures recorded with Range-Doppler radar, we feed voltage excitations directly into our device and detect the skyrmion trajectory evolution. This method scales down to the nanometer level and demonstrates competitive or superior performance compared to energy-intensive software-based neural networks. Our hardware approach’s key advantage is its ability to integrate sensor data in real-time without temporal rescaling, enabling numerous applications.
10 January 2024 in Physical Review Applied
One of the most important magnetic spin structures is the topologically stabilized skyrmion quasiparticle. Its interesting physical properties make it a candidate for memory and efficient neuromorphic computation schemes. For device operation, the detection of the position, shape, and size of skyrmions is required and magnetic imaging is typically employed. A frequently used technique is magneto-optical Kerr microscopy, in which, depending on the sample’s material composition, temperature, material growing procedures, etc., the measurements suffer from noise, low contrast, intensity gradients, or other optical artifacts. Conventional image analysis packages require manual treatment, and a more automatic solution is required. We report a convolutional neural network specifically designed for segmentation problems to detect the position and shape of skyrmions in our measurements. The network is tuned using selected techniques to optimize predictions and, in particular, the number of detected classes is found to govern the performance. The results of this study show that a well-trained network is a viable method of automating data preprocessing in magnetic microscopy. The approach is easily extendable to other spin structures and other magnetic imaging methods.
11 October 2021 in Physical Review Applied
In this work, we explore the stability of magnetic skyrmions confined in a disk geometry by analyzing how to switch a skyrmionic state in a circular disk into a uniformly magnetized state when applying an external magnetic field. The technologically highly relevant energy barrier between the skyrmion state and the uniformly magnetized state is a key parameter needed for lifetime calculations. In an infinite sample, this relates to the out-of-plane rupture field against the skyrmion-core direction, while in confined geometries the topological charge can also be changed by interactions with the sample edges. We find that annihilating a skyrmion with an applied field in the direction of the core magnetization—we call this expulsion—the energy barrier to the uniform state is generally around one order of magnitude lower than the annihilation via the rupture of the core in the disk center, which is observed when the applied field is acting in the direction opposite to the core magnetization. For the latter case a Bloch point (BP) needs to be nucleated to change the topological charge to zero. We find that the former case can be realistically calculated using micromagnetic simulations but that the annihilation via rupture, involving a Bloch point, needs to be calculated with the Heisenberg model because the high magnetization gradients present during the annihilation process cannot be accurately described within the micromagnetic framework.
12 December 2023 in Physical Review Applied
Magnetic skyrmions are topologically stabilized quasiparticles and are promising candidates for energy-efficient applications, such as storage but also logic and sensing devices. Here we present a concept for a multiturn counter-sensor device based on skyrmions, where the number of sensed rotations is encoded in the number of nucleated skyrmions. The skyrmion-boundary force in the confined geometry of the device in combination with the topology-dependent dynamics leads to the effect of automotion for certain geometries. For our case, we describe and investigate this effect with micromagnetic simulations and the Thiele equation in a triangular geometry with an attached reservoir as part of the counter-sensor device. We explore the device functionality, taking into account the influence of thermal diffusion. Additionally, we analyze a readout mechanism for the skyrmion storage, capable of quantifying skyrmion numbers and working effectively even in the presence of thermal diffusion. Finally, our investigation studies the deterministic nucleation of skyrmions.
8 January 2024 in Applied Physics Letters
Magnetic skyrmions are magnetic quasi-particles with enhanced stability and different manipulation mechanisms using external fields and currents, making them promising candidates for future applications such as neuromorphic computing. Recently, several measurements and simulations have shown that thermally activated skyrmions in confined geometries, as they are necessary for device applications, arrange themselves predominantly based on commensurability effects. In this simulational study, based on the Thiele model, we investigate the enhanced dynamics and degenerate non-equilibrium steady state of a system in which the intrinsic skyrmion–skyrmion and skyrmion–boundary interaction compete with thermal fluctuations as well as current-induced spin–orbit torques. The investigated system is a triangular-shaped confinement geometry hosting four skyrmions, where we inject spin-polarized currents between two corners of the structure. We coarse grain the skyrmion states in the system to analyze the intricacies of arrangements of the skyrmion ensemble. In the context of neuromorphic computing, such methods address the key challenge of optimizing readout positions in confined geometries and form the basis for understanding collective skyrmion dynamics in systems with competing interactions on different scales. The findings are corroborated by experimental measurements.