������ý

The race to create superfast computers is accelerating. A rethink of one of the most fundamental parts of a quantum computer could pave the way for ultra-powerful devices.

at the University of New South Wales in Australia and his colleagues have a design for a qubit – the smallest unit of quantum information – that could help get round some of the difficulties of manufacturing quantum computers at an atomic scale.

At the moment, making quantum systems using silicon is difficult because the qubits have to be very close to each other, about 10 to 20 nanometres apart, in order to communicate. This leaves little room to place the electronics needed to make a quantum computer work.

But by combining an electron and nucleus into one qubit, Morello and his team think they’ve found a way to let qubits communicate over distances of up to 500 nanometres. “This would allow you to cram other things between qubits,” says Morello.

Making the leap

Until now, most silicon-based qubits have been made from the electron or the nucleus of a single phosphorus atom. The team’s design uses both the nucleus and the electron of a phosphorus atom to create a single qubit inside a layer of silicon.

Qubits in silicon systems interact through electric fields, and Morello’s team shows that it’s possible to extend the reach of those electric fields by pulling the electron further away from the nucleus of each atom.

This overcomes a couple of the major hurdles that held back silicon-based quantum systems, says at Macquarie University in Sydney, and could eventually make it possible to create quantum computers with millions of qubits that can simulate simple chemical reactions.

Silicon-based qubits aren’t the only candidates for quantum computers. Google is making superconductor-based quantum chips, and claims it is on track to build the first quantum computer capable of surpassing some of the abilities of ordinary computers later this year.

Open race

“Silicon is a bit further behind the pack,” says Devitt. But since the computer industry is already used to building chips out of silicon, silicon is well-placed to catch up or even surpass the performance of other quantum systems. Quantum computers made using silicon qubits might be less-error prone than other systems when it comes to building computers with thousands or millions of qubits, says Devitt.

However, silicon and superconducting quantum systems both only work in temperatures that are close to absolute zero, says at Turing, a quantum start-up in California. She says diamond-based systems could be easier to scale-up because they use similar types of qubits to the silicon systems, but don’t need to be cooled to such extreme temperatures.

“The path is pretty open,” says at RWTH Aachen University in Germany. She says it’s still too early to know which system will end up powering the quantum computers of the future.

We’ll have to wait and see whether this new way of defining a qubit really does unleash the potential of silicon-based quantum computers, says Devitt. “This could be a potential solution that is kind of staring us in the face,” he says. “But they’re going to have to go into the lab and make this work.”

Read more: Quantum simulator with 51 qubits is largest ever