We consider adiabatic superconducting cells operating as an artificial neuron and synapse of a multilayer perceptron (MLP). Their compact circuits contain just one and two Josephson junctions, respectively. While the signal is represented as magnetic flux, the proposed cells are inherently nonlinear and close-to-linear magnetic flux transformers. The neuron is capable of providing the one-shot calculation of sigmoid and hyperbolic tangent activation functions most commonly used in MLP. The synapse features both positive and negative signal transfer coefficients in the range $\sim (-0.5,0.5)$ . We briefly discuss implementation issues and further steps toward the multilayer adiabatic superconducting artificial neural network, which promises to be a compact and the most energy-efficient implementation of MLP.

The imitative modelling of processes in the brain of living beings is an ambitious task. However, advances in the complexity of existing hardware brain models are limited by their low speed and high energy consumption. A superconducting circuit with Josephson junctions closely mimics the neuronal membrane with channels involved in the operation of the sodium-potassium pump. The dynamic processes in such a system are characterised by a duration of picoseconds and an energy level of attojoules. In this work, two superconducting models of a biological neuron are studied. New modes of their operation are identified, including the so-called bursting mode, which plays an important role in biological neural networks. The possibility of switching between different modes in situ is shown, providing the possibility of dynamic control of the system. A synaptic connection that mimics the short-term potentiation of a biological synapse is developed and demonstrated. Finally, the simplest two-neuron chain comprising the proposed bio-inspired components is simulated, and the prospects of superconducting hardware biosimilars are briefly discussed.

The hardware implementation of signal microprocessors based on superconducting technologies seems relevant for a number of niche tasks where performance and energy efficiency are critically important. In this paper, we consider the basic elements for superconducting neural networks on radial basis functions. We examine the static and dynamic activation functions of the proposed neuron. Special attention is paid to tuning the activation functions to a Gaussian form with relatively large amplitude. For the practical implementation of the required tunability, we proposed and investigated heterostructures designed for the implementation of adjustable inductors that consist of superconducting, ferromagnetic, and normal layers.

We present both theoretical and experimental investigations of the proximity effect in a stack-like superconductor/ferromagnetic (S/F) superlattice, where ferromagnetic layers with different thicknesses and coercive fields are made of Co. Calculations based on the Usadel equations allow us to find the conditions at which switching from the parallel to the antiparallel alignment of the neighboring F-layers leads to a significant change of the superconducting order parameter in superconductive thin films. We experimentally study the transport properties of a lithographically patterned Nb/Co multilayer. We observe that the resistive transition of the multilayer structure has multiple steps, which we attribute to the transition of individual superconductive layers with the critical temperature, T_c, depending on the local magnetization orientation of the neighboring F-layers. We argue that such superlattices can be used as tunable kinetic inductors designed for artificial neural networks representing the information in a “current domain”.

High-performance modeling of neurophysiological processes is an urgent task that requires new approaches to information processing. In this context, two- and three-junction superconducting quantum interferometers with Josephson weak links based on gold nanowires are fabricated and investigated experimentally. The studied cells are proposed for the implementation of bio-inspired neurons—high-performance, energy-efficient, and compact elements of neuromorphic processor. The operation modes of an advanced artificial neuron capable of generating the burst firing activation patterns are explored theoretically. A comparison with the Izhikevich mathematical model of biological neurons is carried out.

We present the results of circuit simulations for the adiabatic flux-operating neuron. The proposed cell with a one-shot calculation of activation function is based on a modified single-junction superconducting quantum interferometer. In comparison, functionally equivalent elements of the artificial neural network in the semiconductor-based implementations consist of approximately 20 transistors. Also in the article, we present the connecting synapse based on the adiabatic quantum flux parametron. These neurons and synapses allow constructing artificial neural networks with a magnetic representation of information in the form of direction and/or magnitude of the magnetic flux in the superconducting circuit. We discuss the dissipation of energy during operations in the frame of the proposed concept. This value in superconducting neurons and synapses with sub-nanosecond timescale can be reduced down to 10 aJ and 0.1 aJ respectively. The use of adiabatic superconducting logic circuits in our approach promises compatibility with superconducting quantum information processing systems.