In March 2018, we reported on a paper from Microsoft and its partners in Nature magazine that they had found evidence of the existence of the Majorana quasiparticle. Now, a new preprint has been published on arXiv with a correction indicating the conclusions of the earlier paper were premature. As described in additional detail in an article in WIRED magazine when additional data from the experiment that was not included in the original paper was reviewed, it contradicted the conclusions that the Majorana had been found.

Diagram of a Majorana Quasiparticle Design. Credit: QuTech

Whether or not this Majorana quasiparticle exists or not would be important for the development of a topological qubit. In theory, these types of qubits are not dependent upon a single physical characteristic like spin or energy or polarization, but rather on the overall topology. So as an analogy, from a topological viewpoint a coffee cup and a donut would be considered similar because they both have a hole in the middle. This characteristic makes a topological qubit much more resistant to noise and distortions from the external environment and would reduce the requirements for error correction in a fault tolerant quantum computer. As a hypothetical example, if someone were able to build a 1 million physical qubit superconducting quantum computer, they could user error correction codes that might require a 1000:1 physical-to-logical ratio to end up with 1000 logical qubits. On the other hand, a quantum computer based upon a topological qubit might only require a 10:1 physical-to-logical ratio so that machine would require only 10,000 physical qubits to have 1,000 logical qubits giving it a two orders of magnitude edge over the superconducting implementation. Those sort of numbers would create a tremendous advantage for scaling up and building a fault tolerant quantum processor which is why Microsoft has been investing so heavily in this technology.

To be sure, Microsoft has lots of other quantum activity that is not related to the topological qubit. They recently opened up their Azure Quantum platform for a public preview and support users wishing to use the IonQ and Honeywell processors with others on the way. They have continued developing their software Quantum Development Kit and the Q# programming language, and they recently announced their cryo-CMOS control chips that would be able to support either topological qubits, superconducting qubits, or spin qubits. So they still have a lot of options for being a key participant in the quantum computing market.

Although this setback will probably not end research into the topological qubit, it is clear we won’t see a machine based upon this technology anytime in the near future. In March 2020, well-known quantum physicists Dr. John Preskill and Dr. Jonathan Dowling (since passed away) made a bet on whether a topological quantum two qubit gate would be demonstrated in experimental hardware by March 2030. Preskill bet yes and Dowling bet no. But in a tweet after this recent news broke Preskill wrote “My quote in the article still applies: I am confident we’ll have topological quantum computers at some time in the future, but the time scale is still very uncertain.”

For more about this development, you can read the original 2018 paper titled Quantized Majorana conductance published in Nature magazine, our own report on that original paper published in March 2018, an addendum with an Editorial Expression of Concern published in Nature in April 2020, the new preprint paper that replaces the original one published on arXiv in late January 2021, and articles in WIRED magazine that summarize the latest status change and another one that describes the Preskill/Dowling bet.

February 13, 2021