An expression for the microwave power generated by a spin-torque oscillator in the presence of thermal noise is derived. This expression, when used in a subcritical (I<Ith) regime, demonstrates that generated power is determined by the noise level and provides a simple way to experimentally determine the threshold current Ith for microwave generation. The same expression gives a good quantitative description of the experimentally measured dependence of the generated power on the bias current p(I) in a moderate supercritical (Ith<I<1.5Ith) range of bias current variation.
Time-multiplexed Coherent Ising Machines (CIMs) have demonstrated promising results in rapidly solving large-scale combinatorial problems. However, CIMs remain relatively large and power-demanding. Here, we demonstrate a spinwave-based Ising machine (SWIM) that due to the low spinwave group velocity allows for sufficient miniaturization and reduced power consumption. The SWIM is implemented using a 10-mm-long 5-μm-thick Yttrium Iron Garnet film with off-the-shelf microwave components and can support an 8-spin MAX-CUT problem and solve it in less than 4 μs consuming only 7 μJ. As the SWIM minimizes its energy, we observe that the spin states can demonstrate both uniform and domain-propagation-like switching. The developed SWIM has the potential for substantial further miniaturization with reduction of power consumption, scalability in the number of supported spins, increase of operational speed, and may become a versatile platform for commercially feasible high-performance solvers of combinatorial optimization problems.
We show that in an array of magnetic nanocontacts driven by spin-polarized current a self-phase-locking effect can lead to a regime in which all the contacts are generating microwave oscillations with the same frequency and phase, and coherent addition of microwave power generated by individual nanocontacts is possible. The mechanism of this self-phase-locking effect is strongly nonlinear, while the frequency band of phase locking is an order of magnitude larger than in the usual coupled generators and depends on the direction of the bias magnetic field applied to the nanocontacts.
The temperature dependence of the generation linewidth for an auto-oscillator with a nonlinear frequency shift is calculated. It is shown that the frequency nonlinearity creates a finite correlation time for the phase fluctuations. In the low-temperature limit in which the spectral linewidth is smaller than , the line shape is approximately Lorentzian and the linewidth is linear in temperature. In the opposite high-temperature limit, in which the linewidth is larger than , the nonlinearity leads to an apparent “inhomogeneous broadening” of the line, which becomes Gaussian in shape and has a square-root dependence on temperature. The results are illustrated for the spin-torque nano-oscillator.
It has been shown previously that spin-Hall oscillators based on current-driven bi-layered film structures containing an antiferromagnet (AFM) and a normal metal can generate ultra-short ( 2 ps) “spike-like” pulses in response to an external current stimulus of a sufficient amplitude, thus operating as ultra-fast artificial “neurons.” Here, we report the results of numerical simulations demonstrating that a single AFM “neuron” can perform the logic functions of or, and, majority, or q-gates, while a circuit consisting of a small number of AFM “neurons” can function as a full-adder or as a dynamic memory loop with variable clock frequency. The clock frequencies of such AFM-based logic devices could reach tens of GHz, which make them promising as base elements of future ultra-fast high-efficiency neuromorphic computing.