Industry Innovations

Over the past decade, quantum computing has evolved from a promising field into a race to demonstrate real prototypes. Researchers are experimenting with physical systems including supercooled ions, nitrogen-vacancy centers, and superconducting circuits to serve as qubits. The success of any qubit architecture is contingent on its ability not only to maintain and process delicate quantum states but also to pair with instruments that decipher its output.

Defense contractor Raytheon is on the way to creating a vital cog in that process: an amplifier that can read out the quantum signal from some superconducting qubit systems. Despite its small size, the amplifier has minimal quantum noise. And because it uses microwave technology and is compatible with silicon, it would be easy to manufacture in large quantities using existing infrastructure. “These amplifiers allow the scalable readout of multiple qubits with quantum-limited noise in a footprint that is small enough to integrate into the same package as the qubits themselves,” says Andrew Wagner (pictured above), a lead scientist at Raytheon.

The kinetic inductance traveling-wave amplifiers that Raytheon is working on are not new, but they are a promising contender for early generations of quantum computers. Made from superconducting niobium, they consist of a resonant circuit that sits alongside the qubit circuit—close enough for the two to be coupled, but not so tightly bunched that reading out the amplifier circuit destroys the qubit’s state. “Microwave engineering can become very complicated very quickly the instant you start chaining things together or cramming things into small spaces,” Wagner says.

The amplifier leverages the nonlinearity in a thin layer of niobium nitride to convert the resonant signal of the superconducting circuit to a desired frequency via parametric amplification. To amplify the signal, the phase of the signal and that of a pump wave must match. That matching is achieved by including a band stop or dispersive feature with carefully engineered impedance in the transmission line. “The design of the amplifier requires careful modeling of thin superconducting metal layers in a nontrivial microwave geometry,” Wagner says.

Niobium is a conventional superconductor, with properties that are fully described by the Bardeen-Cooper-Schrieffer (BCS) theory. To simulate the layers, Wagner turned to COMSOL’s finite elements simulator package, which accommodates BCS theory. COMSOL’s package allows Wagner to incorporate the BCS sheet resistance and the superconducting transition temperature so that the kinetic inductance can be accurately calculated.

Wagner says the flexibility of COMSOL’s software is crucial in enabling the precise simulation of the complicated circuits without supercomputing infrastructure. The software maximizes efficiency by having meshing strategies that allow both fine sampling of the solution space around complicated nanoscale components and coarser sampling in less complex, larger-scale areas of the circuit. “That’s important,” Wagner says. “It saves you a ton of memory.”

If Wagner’s designs prove successful, he aims to integrate them into the readout chain of existing qubit circuits. “It’s thrilling to take something you’ve made, cool it down, and make it behave like an atom,” he says.

Phil Dooley is a freelance writer and former laser physicist based in Canberra, Australia.