Most quantum computers are based on superconductors or trapped ions, but an
alternative approach using ordinary atoms may have advantages
A quantum computer that uses ordinary atoms to perform calculations could be
a rival to more exotic devices, although one of its creators says there are
still challenges ahead in scaling up the technology.
The most powerful quantum computers in use today rely on superconductors or
trapped ions to form the basis of their qubits, or quantum bits. Both these
systems have drawbacks: superconducting qubits, like those used by Google,
require ultracold temperatures, while it is hard to arrange trapped-ion or
superconducting qubits so that all of them can communicate with each other.
Now, Mark Saffman at the University of Wisconsin-Madison and his colleagues
have built an alternative quantum computer using six qubits made from
neutrally charged caesium atoms, as opposed to charged ions.
The atoms are trapped in a grid with lasers, spaced far enough away from
each other that they don’t interact. But when individual atoms are excited
by a laser shining at the right frequency, their orbiting electrons move so
far from their parent atoms that they can quantum entangle with their
neighbours – a key phenomenon for a quantum computer.
This two-dimensional structure offers an advantage compared with the set-up
of trapped-ion machines, which are normally configured in a line to avoid
unwanted interactions between the charged particles, limiting their ability
to communicate.
“Because it’s all done with laser beams, you can actually reconfigure the
positions of all your qubits,” says Charles Adams at Durham University, UK,
who wasn’t involved in the work. “So if you decide you want to run a
different algorithm with different connectivity between the qubits, you can
just reprogram where the qubits are.”
Certain algorithms are difficult to run on trapped-ion or superconducting
quantum computers because they require a high amount of connectivity between
qubits. One example is phase-estimation algorithms used in quantum
chemistry, which measure how the state of a quantum system evolves over
time. Such algorithms might prove more feasible on neutral-atom machines.
The team’s device isn’t the first neutral-atom quantum computer, but
previous attempts were designed to model specific physical problems or to
run particular quantum algorithms. Saffman and his colleagues have built the
first fully programmable neutral-atom quantum computer, meaning it can run
any quantum algorithm and could theoretically be scaled up to rival other
leading approaches.
Saffman also works for a company called ColdQuanta that is seeking to
develop a commercial neutral-atom quantum computer. However, he says
considerable obstacles still remain to building larger machines, such as
introducing the ability for qubits to correct errors. “I absolutely don’t
want to overhype where we are. As we progress on developing these machines,
I think the road gets steeper, not the opposite,” he says.
The new device isn’t the only one demonstrating the promise of neutral-atom
machines. French start-up Pasqal has developed a special-purpose
neutral-atom computer with more than 100 qubits, designed for modelling
complex chemistry problems. And Mikhail Lukin at Harvard University and his
colleagues have built a neutral-atom machine that lets qubits entangle with
qubits that are much further than neighbouring ones, though it isn’t fully
programmable.
“These things are now moving into a space where, in the coming years, they
can be serious competition to superconducting qubits and trapped ions,” says
Andrew Daley at the University of Strathclyde, UK. “The rapid development in
the last few years has been exciting.”
Reference: arxiv.org/abs/2112.14589
Tags:
Physics