Yesterday, three members of Microsoft’s quantum team presented their work towards a topological quantum computer at the APS Global Summit in Anaheim. Last month, the team made waves announcing their first topological quantum chip, the Majorana 1. More quietly, Nokia Bell Labs has been working on their own version of a topological quantum computer, and the company claims it’s demonstrated the key ingredients in 2023. Both efforts represent scientific achievements, but bulletproof evidence of a topological quantum bit is elusive.
“I would say all quantum computing is early stages,” says Bertrand Halperin, emeritus professor of physics at Harvard, who is not involved in either effort. “But topological quantum computing is further behind. It could catch up; it’s taking a somewhat different path.”
What’s a Topological Quantum Computer?
Quantum computers run on qubits valued at 0, 1, or some superposition of the two, usually encoded through some local quantum property—say, whether an electron’s spin is up or down. This gives quantum computers different capabilities than their classical cousins, promising to easily crack certain types of problems that are out of reach of even the largest supercomputers. The issue is that these quantum superpositions are very fragile. Any noise in the environment, be it temperature fluctuations or small changes in electric or magnetic fields, can knock qubits out of superposition, causing errors.
Topological quantum computing is a fundamentally different approach to building a qubit, one that in theory would be a much less fragile. The idea is that instead of using some local property to encode the qubit, you would use a global, topological property of a whole sea of electrons. Topology is a field of mathematics that deals with shapes: Two shapes are topologically identical if they can be transformed into each other without tearing new holes or connecting previously unconnected ends. For example, an infinite rope extending into space is topologically distinct from the same rope with a knot in it.
Electrons can “twist” around each other to form something akin to a knot. This knot is more difficult to tie or untie, offering protection against noise. (This is an analogy—the qubits would not be literal knots. For a full technical explanation, see this “short” introduction.)
The issue is that electrons don’t often naturally twist themselves into knots. Theorists have postulated such states could existfor decades, but creating the right conditions for them to arise in practice has been elusive. It’s extremely difficult to make devices that could give rise to knotted electrons, and arguably even more difficult to prove that one has done so.
Microsoft’s “Quantraversy”
The Microsoft team’s approach to creating knotted electrons is to start with a semiconducting nanowire. Then, they layer a superconducting material on top of this nanowire. Both the semiconductor and superconductor layers have to be almost completely devoid of material defects, and held at millikelvin temperatures. In theory, this allows an electron from the semiconducting layer to use the superconductor to effectively spread out over the whole wire, forming something akin to a rope that can be tied into knots. This rope is called a Majorana zero mode.
Definitively showing that they’ve created a Majorana zero mode has proven difficult for the Microsoft team. The team and their collaborators claimed they had achieved this milestone back in 2018, but some researchers were unconvinced by the evidence, saying imperfections in the device could have resulted in the same measurements. The paper got retracted. In 2023, Microsoft and collaborators published further evidence that they’ve created Majoranas, although some scientists have remained unconvinced, and say not enough data was shared to reproduce the results. Last month’s claim remains contentious.
“We are very confident that our devices host Majorana zero modes,” says Chetan Nayak, the lead of the Microsoft effort.
“There is no evidence of even the basic physics of Majoranas in these devices, let alone that you could build a qubit out of them,” says Henry Legg, lecturer at the University of St. Andrews who has authored two preprints disputing Microsoft’s results.
“We would probably all agree that further experiments and better data are necessary before the issue can be considered closed,” Harvard’s Halperin says.
Whether or not the Microsoft team has created Majorana zero modes, making them is just the first step. The team also has to show they can be manipulated to actually do computations. Several types of operations are required to make the kind of knot that represents 0, untie it and tie it into a knot that represents 1, or create a quantum superposition of the two.
The most recent paper demonstrated the team’s capability to do one of the necessary measurements. “It’s a big step,” says Jay Sau, professor of physics at the University of Maryland who has a consulting appointment with the Microsoft team.
In an unusual move, Microsoft’s quantum team held a limited access meeting at their headquarters at Station Q, and invited several researchers in the field. There, they revealed preliminary results demonstrating another such measurement.
“There’s still quite a bit of work to do on that side,” says Michael Eggleston, data and devices leader at Nokia, who was present at the Station Q meeting. “There’s a lot of noise in that system. But I think they’re on a good path.”
To sum up, the Microsoft team has not yet reached the milestone where the scientific community would agree that they’ve created a single topological qubit.
“They have a concept chip which has eight lithographically fabricated qubits,” Eggleston says. “But they’re not functional qubits, that’s the fine print. It’s their concept of what they’re moving towards.”
Nokia Bell Labs quantum computing researchers Hasan Siddiquee (right) and Ian Crawley connecting a dilution refrigerator sample loader for cooldown.Nokia Bell Labs
Nokia’s Approach
A team at Nokia Bell Labs is also pursuing the dream of topological quantum computers, although through a different physical implementation. The team, led by lifelong topological quantum computing devotee Robert Willet, is sandwiching a thin sheet of gallium arsenide in between two other semiconducting slabs. They then cool the sandwich to millikelvin temperaturesand subject it to a strong magnetic field. If the device properties are just right, this could give rise to a two-dimensional version of a global electronic state that can be knotted up. A qubit would require both the creation of this state, and the ability to controllably knot and unknot it.
Robert Willet and his collaborators have also had trouble convincing the scientific community that what they had on their hands are really the highly coveted topological states.
“We’re very confident that we have a topological state,” says Nokia’s Eggleston, who oversees the quantum computing effort.
“I find it reasonably convincing,” Harvard’s Halperin says. “But not everyone would agree.”
The Nokia team has not yet claimed the ability to do operations with the device. Eggleston says they are working on demonstrating these operations, and plan to have results in the second quarter of this year.
Proving Topological Quantum States
Proving the necessary topological ingredients beyond the shadow of a doubt remains elusive. Practically speaking, the most important thing is not whether the exotic topological state can be proven to be present, but whether researchers can build a qubit that is both controllable and much more robust against noise than approaches that are more mature.
Nokia’s team claims that they can maintain error-free quantum superpositions for days, although they cannot control them yet. Data revealed by Microsoft at the Station Q meeting shows their devices remain error-free for 5 microseconds, but they believe this can be improved. (For comparison, a tradition superconducting qubit in IBM’s quantum computer remains error-free for up to 400 microseconds).
“There’s always going to be people who don’t necessarily agree or want more data,” Nokia’s Egglestein says, “and I think that’s the strength of the scientific community to always ask for more. Our feeling on this is you need to scale up complexity of devices.”
“I think at some point you go to the regime where it’s a reasonably good qubit, whether it’s precisely topological or not, that becomes the point of the debate,” Maryland’s Sau says. “But at that point it’s more useful to ask how good or bad of a qubit it is.”
Despite difficulties, topological quantum computing continues to be—at least theoretically—a very promising approach.
“I look at these other qubit types that we see out there today. They’re really nice demonstrations. It’s great science. It’s really hard engineering. Unfortunately, it’s kind of like the vacuum tube back in the 40s,” Egglestein says. “You build computers out of them because that’s all you have, and they’re really challenging to scale up. To me, topological qubits really offer the potential that the transistor did. Something small, something robust, something that’s scalable. And that’s what I think the future of quantum computing is.”
From Your Site Articles
Related Articles Around the Web