A recent study highlights that certain logical operations can be executed with relative ease, while others necessitate adjustments to the error correction methodologies employed. These adjustments are known as lattice surgery and magic state distillation, indicating a lighthearted approach within the field.
In its findings, Microsoft claims to have pinpointed an error correction framework that is notably compact, can be efficiently utilized on hardware capable of housing qubits in photons, atoms, or trapped ions, and supports universal computation. However, Microsoft has yet to demonstrate that this framework functions effectively due to a lack of suitable hardware at their disposal. Currently, Azure provides access to trapped ion machines from IonQ and Qantinuum, which are limited to 56 qubits—significantly short of the 96 required for their preferred iteration of these 4D codes. The most advanced system at their disposal is a 100-qubit machine from PASQAL, which just meets the minimum qubit count but leaves no margin for error.
Although testing smaller versions of these codes is a potential path forward, the Azure team has already shown proficiency with error correction codes utilizing hypercubes, raising questions about the additional benefits of exploring this route.
Future Collaborations
Microsoft appears to be awaiting the development of next-generation machines from its partner, Atom Computing. This upcoming machine is being co-designed with Microsoft and will feature advanced quantum capabilities, as explained by Svore: “This first generation that we are building together between Atom Computing and Microsoft will include state-of-the-art quantum capabilities, with 1,200 physical qubits. The subsequent upgrade will have upwards of 10,000. This progression means we can aim for over a hundred logical qubits, facilitating deeper and more reliable computations.”
