Nature retracted!Microsoft’s massive quantum victory three years ago was a mistake after all

For years, Microsoft has been betting on a type of quantum particle called a Majorana fermion in order to build a quantum computer that actually works. Three years ago, a research team funded by Microsoft published a paper in the journal Nature that proved that such particles do exist, but the discovery was questioned by scholars. Currently, this paper has been withdrawn.

Three years later, Microsoft’s 2018 discovery has proved a failure!

Just now, a research team led by Microsoft has retracted a controversial paper published in the journal Nature in 2018.

The paper is touted as a key breakthrough in the creation of a practical quantum computer, and claims to have found evidence for the elusive subatomic particle known as the Majorana fermion.

An investigation found that the study missed key data.

The paper has now been retracted in the journal Nature.

　Why was it retracted?

The retracted paper comes from a lab led by Microsoft physicist Leo Kouwenhoven at Delft University of Technology.

In March 2018, Leo Kouwenhoven published a paper saying they had observed an elusive particle called a Majorana fermion.

A Majorana fermion is a type of fermion whose antiparticle is itself.

In 1937, Italian physicist Ettore Majorana published a paper positing the existence of such particles, hence the name Majorana fermions.

Compared to Kouwenhoven’s 2012 study, the 2018 paper said it found stronger evidence that Majorana fermions exist.

The publication of this paper has earned Kouwenhoven a great reputation for himself and his laboratory at TU Delft.

The paper claims to have seen the presence of Majorana fermions—“zero-bias peaks”—signals in electrical currents passing through extremely low-temperature semiconductor wires.

Late last month, Kouwenhoven and his 21 co-authors published a new paper that includes more data from their experiments.

The report concluded that they did not find this precious particle after all.

A note from the authors states that the original paper in the prestigious journal Nature will be retracted, citing a “technical error”.

By analyzing raw data from related experiments, two physicists in the field came up with different opinions.

Sergey Frolov, a professor at the University of Pittsburgh, said he sees a lot of problems in the unpublished data, including data points that strayed from the main line but were ignored by the paper.

If you include these data points, the conclusion is completely different – Majorana fermions do not appear.

“We apologize to the community for not having sufficient scientific rigor in the manuscript,” said the paper’s researchers.

　To seize quantum supremacy: Microsoft bets on ‘Majorana fermions’“

In recent years, technology giants such as Google, IBM, Microsoft, and Intel have continued to manipulate more qubits through different technical paths.

Quantum computers are devices made of quantum bits, which can encode data of 1s and 0s, but also use quantum states called superpositions to perform mathematical tricks that bits are impossible to perform in conventional computers.

The main challenge in commercializing this idea is that quantum states are very fragile and can be easily overwhelmed by thermal or electromagnetic noise, making qubits prone to errors.

Google, IBM and Intel have all demonstrated prototype quantum processors of around 50 qubits, and companies including Goldman Sachs and Merck are testing the technology.

But large-scale operational jobs can require thousands of qubits. A quantum computer may need to spend most of its time correcting its own glitches.

As recently as 2004, after Microsoft researchers approached Craig Mundie, head of technology strategy, to say they had a solution to a problem that was holding back quantum computers, the instability of qubits, Microsoft’s response to the Majorana fermion particle got interested.

So Microsoft took a different approach, using papers from theoretical physics to come up with a way to build qubits that it claims would be scalable based on Majorana fermions.

But after more than a decade of research, the research that published the paper turned out to be wrong.

　The dream of quantum computing is broken, and the discovery of Majorana fermions will take 30 years

These elusive Majorana fermions are at the heart of Microsoft’s hardware approach to quantum computing, but that approach is clearly lagging behind other companies such as IBM and Google.

Now, however, the Microsoft team has had to admit that their data interpretation was flawed, dashing Microsoft’s hopes of being the first to launch a working quantum supercomputer.

While Microsoft was looking for Majorana fermions, Google was making steady progress on existing qubit technology.

In 2019, Google developed a 54-qubit computer, Sycamore, which successfully demonstrated “quantum supremacy,” allowing a quantum system to take about 200 seconds to complete a task that would take 10,000 years for a traditional supercomputer to complete.

After that, Microsoft appeared to want to hedge its quantum computing bets, announcing that it would offer other companies access to quantum hardware through its cloud service Azure.

How far Microsoft’s qubits are from that goal is unclear.

University of Maryland theoretical physicist Sankar Das Sarma, who has worked with Microsoft researchers, said that quantum computing based on Majorana fermions could be at a stage similar to the 1926 patent application for the first transistor.

It wasn’t until 1947 that researchers made the first working transistor; the tiny silicon version that enabled the computer industry wasn’t developed until the late 1950s.

“I don’t see why Majorana fermions can’t exist, or they can’t be controlled once they exist,” he said. “But it could take 30 years.”