Evaluation of memory capacity of spin torque oscillator for recurrent neural networks

We have succeeded in experimentally observing the memory functionality of a vortex-type spin torque oscillator used in recurrent neural networks for the first time. Voltage pulses representing random binary pattern of 2000 (1000) bits were applied to the oscillator for learning (test), and the amplitude of the oscillator was evaluated as the 200 virtual nodes. The memory capacity of the whole system including the oscillator was evaluated to be 1.8 at maximum. The value is larger than the memory capacity of the system without the oscillator, indicating the existence of a finite contribution from the spin torque oscillator.

[1]  Harald Haas,et al.  Harnessing Nonlinearity: Predicting Chaotic Systems and Saving Energy in Wireless Communication , 2004, Science.

[2]  K. Mizushima,et al.  Numerical Simulation on Temporal Response of Spin-Torque Oscillator to Magnetic Pulses , 2010 .

[3]  Real-Time Measurement of Temporal Response of a Spin-Torque Oscillator to Magnetic Pulses , 2013 .

[4]  Damien Querlioz,et al.  Spintronic Nanodevices for Bioinspired Computing , 2016, Proceedings of the IEEE.

[5]  S. Urazhdin,et al.  Mutual synchronization of nano-oscillators driven by pure spin current , 2016, 1609.01879.

[6]  A. Fert,et al.  Large microwave generation from current-driven magnetic vortex oscillators in magnetic tunnel junctions. , 2010, Nature communications.

[7]  Damien Querlioz,et al.  Neuromorphic computing with nanoscale spintronic oscillators , 2017, Nature.

[8]  Benjamin Schrauwen,et al.  An experimental unification of reservoir computing methods , 2007, Neural Networks.

[9]  Keisuke Fujii,et al.  Harnessing disordered quantum dynamics for machine learning , 2016, 1602.08159.

[10]  Yoshihiko Horio,et al.  Analogue spin–orbit torque device for artificial-neural-network-based associative memory operation , 2016 .

[11]  Hitoshi Kubota,et al.  Scaling up electrically synchronized spin torque oscillator networks , 2018, Scientific Reports.

[12]  J. Katine,et al.  Mutual phase-locking of microwave spin torque nano-oscillators , 2005, Nature.

[13]  Hitoshi Kubota,et al.  Macromagnetic Simulation for Reservoir Computing Utilizing Spin Dynamics in Magnetic Tunnel Junctions , 2018, Physical Review Applied.

[14]  Yan Zhou,et al.  Perpendicular spin torque promotes synchronization of magnetic tunnel junction based spin torque oscillators , 2009 .

[15]  Johan Åkerman,et al.  Long-range mutual synchronization of spin Hall nano-oscillators , 2016, Nature Physics.

[16]  M. D. Stiles,et al.  Critical current and linewidth reduction in spin-torque nano-oscillators by delayed self-injection , 2015, 1505.04102.

[17]  Vincent Cros,et al.  Efficient Synchronization of Dipolarly Coupled Vortex-Based Spin Transfer Nano-Oscillators , 2015, Scientific Reports.

[18]  Gouhei Tanaka,et al.  Reservoir Computing With Spin Waves Excited in a Garnet Film , 2018, IEEE Access.

[19]  T. Morie,et al.  Self-feedback electrically coupled spin-Hall oscillator array for pattern-matching operation , 2017 .

[20]  A Fukushima,et al.  Spin-torque resonant expulsion of the vortex core for an efficient radiofrequency detection scheme. , 2015, Nature nanotechnology.

[21]  S. Yuasa,et al.  Achievement of high diode sensitivity via spin torque-induced resonant expulsion in vortex magnetic tunnel junction , 2018 .

[22]  Rie Sato,et al.  Frequency transition of spin-torque oscillator under the magnetic-field pulse in nanosecond range , 2011 .

[23]  B. N. Engel,et al.  Phase-locking in double-point-contact spin-transfer devices , 2005, Nature.

[24]  Hitoshi Kubota,et al.  Self-Injection Locking of a Vortex Spin Torque Oscillator by Delayed Feedback , 2016, Scientific reports.

[25]  Henry Markram,et al.  Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations , 2002, Neural Computation.