Microsoft promised a quantum computer by 2029. Nature publishes a review that questions its promise

Microsoft has been trying to demonstrate for almost two decades the existence of Majorana modes: quasiparticles that emerge in certain topological superconducting materials and that behave as if they were Majorana fermions. The Italian physicist Ettore Majorana mathematically described the existence of these fermions in 1937, and since then many researchers have become obsessed with them because they have a characteristic that makes them unique: They are both a particle and their own antiparticle.

However, the Majorana modes that Microsoft works with are not fundamental particles; They are emerging phenomena in the field of condensed matter. Its appeal lies in the fact that if it manages to master them, it could manufacture qubits that are intrinsically more stable than those of its rivals and capable of resisting the external noise that weighs everyone down today. quantum computers.

About that bet Majorana 2 arrived earlier this June, a new topological quantum processor that incorporates new materials to create a more stable topological phase. Its scientific basis comes from an article published in February 2025 in Naturecentral to all subsequent developments of this company. Regarding that work, Chetan Nayak, the technical director of quantum hardware, announced that Microsoft had cut its schedule in half: the goal of having a fully functional quantum computer is now set for 2029.

Shadow looms over Microsoft’s quantum research

The scientific community is being extraordinarily demanding with that article and with the research program surrounding it. In fact, this ability to exercise criticism and analysis is one of the strengths of the scientific method. Last Wednesday, Nature posted a review peer-reviewed and signed by Henry Legg, professor of quantum physics at the University of St. Andrews (Scotland), which raises new doubts about the foundations of that work. Not about a peripheral detail. On its foundations.

And the background of those from Redmond does not help. Two scientific articles supported by Microsoft have already been removed from Nature. On those occasions, the editors also pointed out several alerts about possible problems in two other texts: one from Nature and the other from Science. Microsoft explained that the recalled items were produced outside of its laboratories and that did not review the data before publication. Be that as it may, the February 2025 work questioned by Legg is the fifth under scrutiny and, unlike the previous ones, it is not being withdrawn.

Microsoft’s response has been overwhelming. According to Reutersthis company assures that it supports its research and that its program is making practical advances despite the doubts it is raising. Nayak has illustrated it with a metaphor: “It’s almost like debating whether flight is possible or not, and then you find yourself next to an airplane. Well, come on, get on and take a spin.”

The response has been equally resounding. Sergey Frolov, a physicist at the University of Pittsburgh, points to a fundamental problem: Microsoft lacks the accumulated evidence that supports IBM or Quantinuum, rivals that do not depend on the existence of Majorana modes. “Neither Microsoft nor anyone else has laid the groundwork to make it clear that these advances based on Majorana modes are plausible through a series of reliable experiments,” Frolov declared. “On the contrary, we have several articles that continue to be questioned at the most basic level by different people.”

The geopolitical context adds additional pressure. The Trump Administration has invested $2 billion in quantum computing and this week has set the goal to have a functional system by 2028, a year ahead of Microsoft’s deadline. The race is real. The question is whether the science that supports it is also.

Image | Microsoft

More information | Reuters

In Xataka | 38% of AI experts in the US have been trained in China. They are essential to sustain your leadership