A fluxonium qubit can keep its most useful quantum properties for about 1.48 milliseconds, drastically longer than similar qubits currently favoured by the quantum computing industry.
A superconducting qubit, or quantum bit, has broken the record for how long it can maintain its quantum properties. Extending this time will make future quantum computers more useful.
 |
| Fluxonium qubits could make quantum computers more useful Bartlomiej K. Wroblewski/Alamy |
The first step in building a quantum computer is choosing how to make its key ingredients, called qubits. One popular choice, championed by research labs and industry players such as IBM and Google, is the transmon superconducting qubit. But like all qubits, these can become ineffective at storing and processing information after a short time if there are any small disturbances in their environment.
Aaron Somoroff at the University of Maryland and his colleagues have now showed that a fluxonium qubit, a more complex superconducting cousin of the transmon, could stay useful for longer.
They made their fluxonium qubit by laying extremely thin wires of titanium and aluminium onto a sapphire chip in a special configuration, creating many channels between rows of superconducting “islands”. These wires only superconduct when they are extremely cold, so the researchers kept the chip in a refrigerator at a temperature of about 0.008 kelvin.
When they ran electricity through the chip, the special layout and superconductivity of the wires made it have several distinct quantum states. Each of these could be used to encode information into 1s and 0s, or “on” and “off”, similar to how ordinary computers work. Because these were quantum states, however, the chip also offers an infinite amount of options between 1 and 0. To assess how useful these many states are, the team measured the chip’s coherence time, which shows how long a qubit can store information without error.
Somoroff says that the best transmon qubits have coherence times of hundreds of microseconds, but he and his team measured about 1.48 milliseconds for their fluxonium qubit. They also determined that they could change their qubit’s state, something that would have to happen many times during a computation on a fluxonium quantum computer, with 99.991 per cent fidelity. This makes the fluxonium qubit one of the most reliable qubits that exists, almost always changing states exactly as instructed.
These properties are an important step towards more practical quantum computers, says Chen Wang at the University of Massachusetts Amherst. But there are still more to develop.
“There are many things to consider in building a useful quantum computer. You do need a long coherence time and to operate qubits at fast speeds. But you also need to be able to read out information without errors and have multiple qubits talk to each other,” he says.
“A coherence time in milliseconds catches my eye,” says Ian Mondragon Shem at Northwestern University in Illinois. “Improving coherence times could increase the chances of fluxonium hardware becoming more common.”
Mondragon Shem says that the superconducting circuits that comprise fluxonium qubits are more complicated and could be more challenging to manufacture than those that are routinely made for transmons. Regardless, he says that fluxonium has been gaining traction among researchers in recent years.
For his part, Somoroff isn’t intimidated by possible manufacturing issues. “We were able to fabricate fluxonium qubits pretty reliability in an academic setting with a small team, so a company like Google or IBM could do it, too. I actually think that we’re getting pretty close to fluxonium overtaking the transmon,” he says.
Journal reference:
Physical Review Letters,