Microsoft quantum computing 'breakthrough' faces fresh challenge

Analysis pokes holes in protocol that underpins Microsoft’s claim to have created the first topological qubits

Microsoft unveiled its 'Majorana 1' quantum chip in February.Credit: John Brecher for Microsoft

A physicist has cast doubt on a test that underlies a high-profile claim by Microsoft to have created the first ‘topological qubits’, a long-sought goal of the company’s quantum computing effort. The critique comes amid mounting speculation about the validity of Microsoft’s claim.

Microsoft announced the breakthrough, which could lead to a quantum computer more resistant to information loss than with other approaches, on 19 February. Without a peer-reviewed paper backing up the claim, some researchers were sceptical. An accompanying paper in Nature described a method to measure the read-out from future topological qubits, but did not offer proof of their existence¹.

“While the Nature paper outlined our approach, it does not speak to our progress,” a Microsoft spokesperson said in a statement. The paper was submitted almost a year before it was published and since then “tremendous progress has occurred”, they said. (Nature’s news team is independent of its journals team.)

In the latest critique, posted as a preprint², Henry Legg, a theoretical physicist at the University of St Andrews, UK, raises concerns about a test that Microsoft uses to look for Majoranas, so-far undiscovered quasiparticles arising from the collective behaviour of electrons that are needed for the topological qubits to work.

Known as the topological gap protocol (TGP), the test is not mentioned in the 19 February Microsoft announcement. But the company has subsequently indicated to Nature’s news team, and in a comment online, that it created the topological qubits using the TGP. “Since the TGP is flawed, the very foundations of the qubit are not there,” says Legg.

Chetan Nayak, a theoretical physicist leading Microsoft’s quantum computing effort in Redmond, Washington, stands by the qubit claim. “The criticism can be summarized as Legg constructing a false straw man of our paper and then attacking that,” he says.

A Majorana test

Majoranas have previously proved elusive. Several claimed sightings later proved to be Majorana mimics, in some cases leading to retractions.

In 2022, Microsoft researchers posted a preprint reporting that the TGP could indirectly identify Majoranas with ‘high probability’; the preprint was later published in Physical Review B (PRB)³. The test relies on electronic measurements of microscopic sandwiches of ultracold metal. If the measurement exhibits a specific feature characteristic of Majoranas, the device ‘passes’ and is said to host the elusive particles.

Legg and his colleagues at the University of Basel in Switzerland went on to report that the test could be fooled by false positives, doppelgangers with the electronic features of Majoranas that lacked their useful properties⁴.

In his latest critique, Legg reports further flaws with the protocol. Examining data in the PRB paper, he saw large variations in the external conditions, such as the range of magnetic field strengths, when electronic measurements were taken. The test was designed to measure the intrinsic properties of the device — and thus whether it contained Majoranas. Instead, Legg found that varying conditions made the test an inconsistent weathervane.

Nayak does not accept this criticism. “The ranges come from an initial scan we describe, and we always analyze the full data,” he says.

Another problem, says Legg, is that a key parameter in the code that implements the protocol, which Microsoft shared publicly, differs from the description in the PRB paper. When asked by Nature’s news team about the issue, Nayak said: “Legg claims there’s a difference between our described protocol and the implemented code. This is incorrect, so this is a non-issue.” Legg says that Nayak previously acknowledged this difference in an e-mail to him and planned to issue a correction, but changed his mind.

Legg’s critique is “certainly valid”, says Carlo Beenakker, a theoretical physicist at the University of Leiden in the Netherlands, although he remains “enthusiastic” about Microsoft’s goal of creating topological qubits. “I have a deep respect for people who basically go out and somehow are not distracted by others who say, ‘I don't believe it,’” he says.

Anton Akhmerov, a theoretical physicist at Delft University of Technology in the Netherlands, says that “Legg's critique requires a public response from the Microsoft researchers”. A Microsoft spokesperson said that they would officially respond to Legg's critique when contacted by editors of PRB.