Accelerating Computing with Light
Avik Dutt is using quantum ingenuity to sidestep the limits of conventional computer chips—all while using commercially accessible components.
Quantum Leap
For decades, computers have steadily grown faster and more powerful. Avik Dutt, an assistant professor of mechanical engineering, envisions another dramatic leap forward—one propelled by the quantum properties of light.
“With light-based quantum computing, we could solve a problem in minutes or even seconds that, today, would take hundreds of years,” he says.
In the future, quantum computers could crack the complex problems with which today’s computers struggle—for example, designing new drugs and new materials, and improving weather forecasts and navigation apps.
A brief guide to quantum strangeness
- Entanglement: A quantum link between two or more particles—even at a distance—that correlates them.
- Qubit: Short for quantum bit. The basic unit of information in quantum computing. Distinct from a classical bit, which must have a value of one or zero.
- Superposition: A quantum particle’s ability to exist in a combination of states simultaneously. It allows qubits to have a value of neither zero nor one, but some probability of either at once.
Size comparison: A dime versus Dutt’s quantum chip
Light-carrying fiber optic cables already form the backbone of modern global communication networks. But conventional fiber optics rely on the classical behavior of light; Dutt works at the more fundamental quantum scale, where different rules apply.
Quantum’s light advantage
Light isn’t the only potential medium for quantum computing; researchers are also exploring other possibilities, such as superconducting materials and trapped ions.
Dutt favors light because, unlike these alternatives, it does a superior job of maintaining quantum properties, such as superposition and entanglement, in commonplace room-temperature environments without needing vacuum.
To manipulate packets of light (called photons) at the quantum level, his lab designs chips much like those currently used in computer processors. This approach offers a distinct advantage: It means that the same nanoscale fabrication facilities that make the electronic chips in cell phones, cars, and so many other devices could also produce light-based quantum technology.
Rather than waiting on the construction of new factories, this new field could instead draw on existing manufacturing capacity—potentially speeding the day when quantum tools reach the market at an affordable cost.
Speeding up sensing with quantum light
Dutt’s light-based quantum approach could also speed up another technology: sensors.
His team has built such a device to detect the toxic gas hydrogen cyanide, using light manipulated by ring resonators on a silicon nitride chip. Changes in the light’s quantum properties as it passes through the chip make it possible to sense the gas at concentrations invisible to conventional instruments.
Like other quantum sensors—such as those intended for navigation, or for noninvasive monitoring of molecules in the body—Dutt’s detector works much faster, potentially providing critical information in milliseconds.
Encode more info with simpler hardware
Quantum systems, however, still need further development before they are ready for mass production. Dutt’s lab recently tackled a significant challenge they face: The need to scale up.
For conventional electronics, developers have historically increased information processing speed and power by cramming more, ever-tinier transistors onto chips. In research published recently in the journal Nature Physics, Dutt’s team demonstrated physics that points toward a different approach.
“Instead of trying to create complexity by making more and more repeating physical structures, we wanted to use internal control to simulate a large number of dimensions in a simpler system,” says Sashank Sridhar, a graduate student in Dutt’s lab.
Each added dimension, he explains, is an extra resource that allows the same physical structure to transmit more information. Rather than encoding one quantum bit (qubit) as a single photon in a physical wire, researchers could encode many qubits across several frequency modes of the same component: “Shuttling light through these synthetic dimensions would allow us to carry and process the equivalent of many qubits within one physical structure, all on a compact, chip-scale device.”
Dutt designs quantum chips with off-the-shelf components.
Quantum computing’s new physics, from off-the-shelf components
To demonstrate the underlying physics of its light-based quantum computing, the team built two ring-shaped optical resonators using commercially available fiber optics and coupled them. By applying modulation signals at carefully chosen microwave frequencies, they drove light energy back and forth between the two rings. Each added frequency contributed a new synthetic dimension, encoded in the internal properties of the light rather than in physical space.
Notably, the team showed they could generate synthetic dimensions under real-world conditions, and they proposed a chip-based platform for achieving the same result.
“It’s quite remarkable that you can do fundamentally new physics with just ingenious engineering,” Sridhar says.
Top