The article therefore describes how some logical operations can be performed relatively easily, while a few others require manipulations of the error correction schedule to work. (These manipulations have names such as scheduling surgery and magical state shilling, which are good signs that the field does not take itself That serious.)
So, together, Microsoft is of the opinion that it has identified a error correction schedule that is fairly compact, can be efficiently implemented on hardware that stores quubits in photons, atoms or imprisoned ions and universal calculation. What it did not do, however, has shown that it really works. And that's because it just doesn't have the hardware at the moment. Azure offers trapped ion machines from IONQ and Qantinuum, but this top at 56 Qubits – well under the 96 needed for their favorite version of these 4D codes. The biggest thing it has access is a machine of 100 breeding from a company called Pasqal, which hardly matches the 96 quubits that are needed, so there is no room for errors.
Although it should be possible to test smaller versions of codes in the same family, the Azure team has already shown that the ability to work with error correction codes based on hypercubes, so it is unclear whether there is anything to be gained in that approach.
More atoms
Instead, it seems to be waiting for another partner, Atom Computing, to place its next generation machine, one that it designs in collaboration with Microsoft. “This first generation that we build together between Atom Computing and Microsoft will contain ultramodern quantum options, will have 1,200 physical Qubits,” said Svore “and then the next upgrade of that machine will have more than 10,000. And so you look at more than a hundred logical Qubits with deeper and more reliable.”
