Machine learning-powered compact modeling of stochastic electronic devices using mixture density networks.

The relentless pursuit of miniaturization and performance enhancement in electronic devices has led to a fundamental challenge in the field of circuit design and simulation-how to accurately account for the inherent stochastic nature of certain devices. While conventional deterministic models have served as indispensable tools for circuit designers, they fall short when it comes to capturing the subtle yet critical variability exhibited by many electronic components. In this paper, we present an innovative approach that transcends the limitations of traditional modeling techniques by harnessing the power of machine learning, specifically Mixture Density Networks (MDNs), to faithfully represent and simulate the stochastic behavior of electronic devices.

All optical operation of a superconducting photonic interface.

Quantum photonic processing via electro-optic components typically requires electronic links across different operation environments, especially when interfacing cryogenic components such as superconducting single photon detectors with room-temperature control and readout electronics. However, readout and driving electronics can introduce detrimental parasitic effects. Here we show an all-optical control and readout of a superconducting nanowire single photon detector (SNSPD), completely electrically decoupled from room temperature electronics.

Photonic online learning: a perspective.

Emerging neuromorphic hardware promises to solve certain problems faster and with higher energy efficiency than traditional computing by using physical processes that take place at the device level as the computational primitives in neural networks. While initial results in photonic neuromorphic hardware are very promising, such hardware requires programming or "training" that is often power-hungry and time-consuming. In this article, we examine the online learning paradigm, where the machinery for training is built deeply into the hardware itself.

Layer-Skipping Connections Improve the Effectiveness of Equilibrium Propagation on Layered Networks.

Equilibrium propagation is a learning framework that marks a step forward in the search for a biologically-plausible implementation of deep learning, and could be implemented efficiently in neuromorphic hardware. Previous applications of this framework to layered networks encountered a vanishing gradient problem that has not yet been solved in a simple, biologically-plausible way. In this paper, we demonstrate that the vanishing gradient problem can be mitigated by replacing some of a layered network's connections with random layer-skipping connections in a manner inspired by small-world networks.

A Stochastic SPICE Model for Superconducting Nanowire Single Photon Detectors and Other Nanowire Devices.

Superconducting nanowire devices, such as the superconducting nanowire single photon detector (SNSPD) or nanocryotron, have a time-dependent stochasticity that depends on the current flowing through them. When modeling complex circuits made of several such devices (for instance, an array of SNSPDs), the ability to include this randomness can be important for predicting unwanted effects and interactions within the circuit. We present a modification of the model described by Berggren .

A kinetic-inductance-based superconducting memory element with shunting and sub-nanosecond write times.

We present a kinetic-inductance-based superconducting memory element with non-destructive readout, femtojoule read and write energies, both read and write shunts, which is writeable with pulses shorter than 400 ps. The element utilizes both a high-kinetic-inductance layer made from tungsten silicide as well as a low-kinetic-inductance layer made from niobium. By using tungsten silicide-which has a long (20 ns) thermal time constant-and measuring bit error rates from 10 MHz to 1 GHz, we were able to verify that the thin-film elements could be operated at a data rate at least as fast as the material thermal time constant with a bit error ratio less than 10.

All-silicon light-emitting diodes waveguide-integrated with superconducting single-photon detectors.

We demonstrate cryogenic, electrically injected, waveguide-coupled Si light-emitting diodes (LEDs) operating at 1.22 m. The active region of the LED consists of W centers implanted in the intrinsic region of a -- diode.

Using Geometry To Sense Current.

We describe a superconducting three-terminal device that uses a simple geometric effect known as current crowding to sense the flow of current and actuate a readout signal. The device consists of a "Y"-shaped current combiner, with two currents (sense and bias) entering separately through the top arms of the "Y", intersecting, and then exiting together through the bottom leg of the "Y". When current is added to or removed from one of the arms (e.

nanoSQUID operation using kinetic rather than magnetic induction.

We report on a method of nanoSQUID modulation which uses kinetic inductance rather than magnetic inductance to manip-ulate the internal fluxoid state. We produced modulation using injected current rather than an applied magnetic field. Using this injected current, we were able to observe the triangle-wave shaped modulation of the device critical current which was periodic according to the London fluxoid quantization condition.

Free-space-coupled superconducting nanowire single-photon detectors for infrared optical communications.

This paper describes the construction of a cryostat and an optical system with a free-space coupling efficiency of 56.5% ± 3.4% to a superconducting nanowire single-photon detector (SNSPD) for infrared quantum communication and spectrum analysis.

Eight-fold signal amplification of a superconducting nanowire single-photon detector using a multiple-avalanche architecture.

Superconducting nanowire avalanche single-photon detectors (SNAPs) with n parallel nanowires are advantageous over single-nanowire detectors because their output signal amplitude scales linearly with n. However, the SNAP architecture has not been viably demonstrated for n > 4. To increase n for larger signal amplification, we designed a multi-stage, successive-avalanche architecture which used nanowires, connected via choke inductors in a binary-tree layout.

A superconducting-nanowire three-terminal electrothermal device.

Superconducting electronics based on Josephson junctions are used to sense and process electronic signals with minimal loss; however, they are ultrasensitive to magnetic fields, limited in their amplification capabilities, and difficult to manufacture. We have developed a 3-terminal, nanowire-based superconducting electrothermal device which has no Josephson junctions. This device, which we call the nanocryotron, can be patterned from a single thin film of superconducting material with conventional electron-beam lithography.