Reservoir computing with the frequency, phase, and amplitude of spin-torque nano-oscillators

@article{Markovic2019ReservoirCW,
  title={Reservoir computing with the frequency, phase, and amplitude of spin-torque nano-oscillators},
  author={Draza Markovic and Nathan Leroux and Mathieu Riou and Flavio Abreu Araujo and Jacob Torrejon and D. Querlioz and Akio Fukushima and Shutaro Yuasa and Juan Trastoy and Paolo Bortolotti and Julie Grollier},
  journal={Applied Physics Letters},
  year={2019}
}
Spin-torque nano-oscillators can emulate neurons at the nanoscale. Recent works show that the non-linearity of their oscillation amplitude can be leveraged to achieve waveform classification for an input signal encoded in the amplitude of the input voltage. Here we show that the frequency and the phase of the oscillator can also be used to recognize waveforms. For this purpose, we phase-lock the oscillator to the input waveform, which carries information in its modulated frequency. In this way… 

Figures from this paper

Temporal pattern recognition with delayed feedback spin-torque nano-oscillators.

TLDR
This study extends the memory of the spin-torque nano-oscillators through time-delayed feedback to increase the efficiency of solving pattern recognition tasks that require memory to discriminate different inputs.

Radio-Frequency Multiply-and-Accumulate Operations with Spintronic Synapses

TLDR
This work demonstrates through physical simulations with parameters extracted from exper-imental devices that frequency-multiplexed assemblies of resonators implement the corner-stone operation of artificial neural networks, the Multiply-And-Accumulate (MAC), directly on microwave inputs.

Non-volatile artificial synapse based on a vortex nano-oscillator

TLDR
By using two non-identical STVOs, it is shown how these devices can be used as programmable non-volatile synapses in artificial neural networks.

Giant magnetoresistance amplifier for spin-orbit torque nano-oscillators

Spin-orbit torque nano-oscillators based on bilayers of ferromagnetic (FM) and nonmagnetic (NM) metals are ultra-compact current-controlled microwave signal sources. They serve as a convenient

Implementing a magnonic time-delay reservoir computer model

Recently we demonstrated experimentally that microwave oscillators based on the time delay feedback provided by traveling spin waves could operate as reservoir computers. In the present paper, we

Spin–orbit torque nano-oscillator with giant magnetoresistance readout

Spin-orbit torque nano-oscillators based on bilayers of ferromagnetic and nonmagnetic metals are ultra-compact current-controlled microwave signal sources. They are attractive for practical

Focused surface acoustic wave induced nano-oscillator based reservoir computing

We demonstrate using micromagnetic simulations that a nanomagnet array excited by surface acoustic waves (SAWs) can work as a reservoir. An input nanomagnet is excited with focused SAW and coupled to

Bifurcation analysis of strongly nonlinear injection locked spin torque oscillators

. We investigate the dynamics of an injection locked in-plane uniform spin torque oscillator for several forcing configurations at large driving amplitudes. For the analysis, the spin wave amplitude

Easy-plane spin Hall nano-oscillators as spiking neurons for neuromorphic computing

We show analytically using a macrospin approximation that easy-plane spin Hall nano-oscillators excited by a spin-current polarized perpendicularly to the easy-plane have phase dynamics analogous to

Reservoir Computing on Spin-Torque Oscillator Array

TLDR
It is shown that reservoir computing can be performed using the synchronized oscillation state of the spin-torque oscillator (STO) array, and STO arrays are promising for hardware implementation of reservoir computing.
...

References

SHOWING 1-10 OF 28 REFERENCES

Neuromorphic computing with nanoscale spintronic oscillators

TLDR
It is shown experimentally that a nanoscale spintronic oscillator can be used to achieve spoken-digit recognition with an accuracy similar to that of state-of-the-art neural networks and the regime of magnetization dynamics that leads to the greatest performance is determined.

Vowel recognition with four coupled spin-torque nano-oscillators

TLDR
The results demonstrate that non-trivial pattern classification tasks can be achieved with small hardware neural networks by endowing them with nonlinear dynamical features such as oscillations and synchronization, and that the high experimental recognition rates stem from the ability of these oscillators to synchronize.

Coupled Spin Torque Nano Oscillators for Low Power Neural Computation

TLDR
Results show that injection locking-based neuron model can be attractive from scaling point of view and computation like character recognition can be performed with energy consumption per neuron of ~1.8× and ~ 3× lower than the digital and the analog CMOS counterpart, respectively.

Nonwhite frequency noise in spin torque oscillators and its effect on spectral linewidth

We measure the power spectral density of frequency fluctuations in nanocontact spin torque oscillators over time scales up to 50 ms. We use a mixer to convert oscillator signals ranging from 10 GHz

Time Required to Injection-Lock Spin Torque Nanoscale Oscillators

We have injection-locked a spin-transfer oscillator to a second-harmonic electrical input signal and measured the relative phase and amplitude of the device output as a function of DC current under

Nonlinear Auto-Oscillator Theory of Microwave Generation by Spin-Polarized Current

This paper formulates a general analytic approach to the theory of microwave generation in magnetic nano-structures driven by spin-polarized current and reviews analytic results obtained in this

Injection locking and phase control of spin transfer nano-oscillators.

TLDR
Time-domain measurements show that the phase of the spin transfer oscillations varies over a range of approximately +/-90 degrees relative to the input, in good agreement with general theoretical analysis of injection locking of nonlinear oscillators.

Amplitude and phase noise of magnetic tunnel junction oscillators

The microwave emission linewidth of spin transfer torque nano-oscillators is closely related to their phase and amplitude noise that can be extracted from the magnetoresistive voltage signal V(t)

Spin-wave interference patterns created by spin-torque nano-oscillators for memory and computation

TLDR
This work shows how arrays of spin-torque nano-oscillators can create propagating spin-wave interference patterns of use for memory and computation and groups of transponders create resonant (reverberating) spin- wave interference patterns that may be used for polychronous wave computation and information storage.

Spin-wave-beam driven synchronization of nanocontact spin-torque oscillators.

TLDR
Not only is robust synchronization between two oscillators observed for separations larger than 1,000 nm, but synchronization of up to five oscillators, a new record, has been observed in the vertical array geometry.