An expert unpacks the hype about Microsoft’s new topological qubit approach
“If people think they’re going to be solving really important problems on topological quantum computers in five years, I think that’s very unlikely,” says Peter Love. Here, a Majorana 1 chip. Photo: John Brecher/Microsoft
Quantum computing is the next big step for computers—and it’s still a long way off. It holds the promise of machines that will be able to solve immensely difficult problems, such as modeling properties of molecules or breaking cryptography systems.
Major research efforts are underway to build different types of quantum computers, which go beyond the binary 1s and 0s of computers we currently use, instead being able to hold states of both 1 and 0 and quantum states in between the two, adding exponential computing power for certain problems. There are different types of quantum computers already under development—those using superconducting materials, ion traps, and photons.
Now Microsoft has announced what it calls “a breakthrough class of materials called a topoconductor … [that] marks a transformative leap toward practical quantum computing.”
“This is really exciting physics, and it’s great that they’re pursuing it with this level of intensity,” says Peter Love, a professor of physics who has worked in the quantum computing field for several decades. “But if people think they’re going to be solving really important problems on topological quantum computers in five years, I think that’s very unlikely.”
In quantum computing, the fundamental unit is a qubit, short for quantum bit. The new Microsoft qubits design uses topology—a branch of mathematics concerned with shapes and forms—it might perform better than those currently being developed.
Quantum computers now in existence all face a common problem: “noise” that introduces errors in the computation. The noise stems from the extremely sensitive nature of the qubits in a quantum state.
Superconducting qubits are fixed onto a substrate, Love says, and invariably interact at the atomic level with the substrate material, leading to aberrations in the signal. Such errors have to be corrected for.
The advantage of the topological qubits that Microsoft is working on is that “they offer more resilience to noise, and therefore, one hopes, lower error rates,” he says.
Is It or Isn’t It?
Microsoft detailed how the topological qubit works in a research paper published in the journal Nature, saying it generated what’s called a Majorana mode. Named for an early 20th century Italian physicist, this mode is key to the qubit’s function. The journal made peer reviewer reports available, and some of those argue that this isn’t the only interpretation of what the researchers saw.
“The other interpretation more or less is that this is not a Majorana state, rather something called an Andreev bound state,” says Love. “The upshot is that it could be something else, and this experiment can’t distinguish between those bound states and the Majorana states. Microsoft, of course, is betting that it’s a Majorana state, but the possibility that it’s not is still there.”
How to know for sure? He points out that while current quantum computers have up to several hundred qubits, the Microsoft team announced its discovery after making just one qubit. More examples would help.
“If Microsoft makes two or eight of these things, and they keep behaving like qubits, and you can see topological states, then that’s probably the best way of settling the question,” he says.
Microsoft called its proof-of-concept topological qubit “the world’s first Quantum Processing Unit (QPU) powered by a topological core, designed to scale to a million qubits on a single chip.” In the media hype that immediately followed the announcement, some took that million-qubit number as a done deal. Love offers caution.
“What they mean is that they’ve made one in such a way that they think they can then stamp them out—so good luck with that, right? Tell me when you’ve got two or 10 some years from now,” Love says. “They’re talking about a million qubits, but you have to understand that they don’t have that. Still, it does matter that they’re thinking architecturally.”
That said, “the rate of production of qubits and the rate of progress on the experimental side has always been greater than I anticipated,” he says. “We’ve got a long way to go to have a useful application, but we’re a lot further ahead than I ever thought we would be.”
The announcement spurred a large amount of interest in quantum computing, and Love says he’s happy for that. “If you’re interested in quantum computing, there’s another layer of something to understand all the time, and hopefully this is an opportunity for people to do that,” he says.
“Everything in quantum computing is very challenging, which is partly why it’s an exciting field, pushing the envelope of what you can do with physics, which is fantastic,” he says.
