Journal of Applied Physics · 2018 · 32 citations · 14 references
Life SciencesBioengineeringNeuroscienceCellular NeurobiologyNeural CircuitryNeural EngineeringMembrane ExcitabilityNeural ExcitabilityNeuronal DynamicsSpintronicsSynaptic SignalingSynaptic TransmissionDwell-time AnalysisMagnetic Tunnel JunctionsSignalsSpike Train DecodingSpike TrainsSpiking ActivityTuningTunnel Junctions
Magnetic tunnel junctions (MTJs) operating in the superparamagnetic regime produce telegraphic signals that emulate neural spiking signals. Previous studies have characterized the random spiking signals produced by MTJs in terms of the percentage of time spent in the anti-parallel (AP) magnetization state (referred to as the “AP rate”) but ignore the switching rate of the MTJ. In this work, we demonstrate that with proper tuning of both an external bias field and a bias voltage, we can control the average dwell time in the AP-state and P-state pulses separately. Our data show that the AP rate can be tuned with bias voltages ranging from 310 mV to 460 mV and bias fields from −200 Oe to −230 Oe. The average dwell times in each state ranged from 225 ns to 285 μs and could be controlled separately. This suggests that neural spiking signals produced by MTJs can be decoded by both the spike rate and the spike count, which creates the possibility for increasing the information capacity in the rate coding scheme.
14
Hui Zhao, Andrew Lyle, Y. Zhang et al. · Journal of Applied Physics · 2011 · 123 citations
Natural Sciences, Materials Engineering, Materials Science +16
Low-Barrier Nanomagnets as p-Bits for Spin Logic
Rafatul Faria, Supriyo Datta · IEEE Magnetics Letters · 2017 · 85 citations · Full text