Step-like dependence of memory function on pulse width in spintronics reservoir computing

Abstract

Physical reservoir computing is a type of recurrent neural network that applies the dynamical response from physical systems to information processing. However, the relation between computation performance and physical parameters/phenomena still remains unclear. This study reports our progress regarding the role of current-dependent magnetic damping in the computational performance of reservoir computing. The current-dependent relaxation dynamics of a magnetic vortex core results in an asymmetric memory function with respect to binary inputs. A fast relaxation caused by a large input leads to a fast fading of the input memory, whereas a slow relaxation by a small input enables the reservoir to keep the input memory for a relatively long time. As a result, a step-like dependence is found for the short-term memory and parity-check capacities on the pulse width of input data, where the capacities remain at 1.5 for a certain range of the pulse width, and drop to 1.0 for a long pulse-width limit. Both analytical and numerical analyses clarify that the step-like behavior can be attributed to the current-dependent relaxation time of the vortex core to a limit-cycle state. }

9 pages, 5 figures

References in corpus (16)

• Microwave Oscillations of a Nanomagnet Driven by a Spin-Polarized Current
• Current-driven Magnetization Reversal and Spin Wave Excitations in Co/Cu/Co Pillars
• Direct-Current Induced Dynamics in Co90Fe10/Ni80Fe20 Point Contacts
• Physical reservoir computing -- An introductory perspective
• Harnessing disordered quantum dynamics for machine learning
• Magnetic Vortex Core Dynamics in a Ferromagnetic Dot
• Potential implementation of Reservoir Computing models based on magnetic skyrmions
• Vortex oscillations induced by a spin-polarized current in a magnetic nanopillar: Evidence for a failure of the Thiele approach
• Boosting computational power through spatial multiplexing in quantum reservoir computing
• Macromagnetic simulation for reservoir computing utilizing spin dynamics in magnetic tunnel junctions
• Reservoir computing with the frequency, phase and amplitude of spin-torque nano-oscillators
• Low-frequency vortex dynamic susceptibility and relaxation in mesoscopic ferromagnetic dots
• Temporal pattern recognition with delayed feedback spin-torque nano-oscillators
• Reservoir computing with dipole-coupled nanomagnets
• Reservoir computing with simple oscillators: Virtual and real networks
• Periodic structure of memory function in spintronics reservoir with feedback current

Cited by in corpus (10)

• High Speed Photonic Neuromorphic Computing Using Recurrent Optical Spectrum Slicing Neural Networks
• Spintronic reservoir computing without driving current or magnetic field
• Handwritten Digit Recognition by Spin Waves in a Skyrmion Reservoir
• Computational capability for physical reservoir computing using a spin-torque oscillator with two free layers
• Input-driven chaotic dynamics in vortex spin-torque oscillator
• Non-periodic input-driven magnetization dynamics in voltage-controlled parametric oscillator
• Magneto-Ionic Physical Reservoir Computing
• Predicting the future with magnons
• Learning-Performance Evaluation of a Physical Reservoir Based on a Vortex Spin-Torque Oscillator with a Modified Free Layer
• Focused Surface Acoustic Wave induced nano-oscillator based reservoir computing