MIT and Harvard team designs compact error correction codes for atom arrays

[Andy89]

Hengyun Zhou's group at the MIT Harvard Center for Ultracold Atoms has been working to reduce the overhead of error correction on neutral atom quantum computers. Quantum Zeitgeist's Ivy Delaney has covered a research highlight that pulls together several of the group's papers. The common theme is doing more with fewer atoms

The team designed codes for reconfigurable atom arrays that store more than one logical qubit for every two physical data qubits. That is very efficient by error correction standards, where many physical qubits are usually needed for each logical one. Simulations showed very low error rates, even when errors in the checking circuits themselves were included

A follow up with Jong Yeon Lee set out design rules that balance storage, protection and the complexity of checks. That produced families of codes built in blocks of only a few hundred physical qubits. The codes have a distance of 12, which gives strong protection

Other work used the codes' symmetries. A study led by Nishad Maskara with Mikhail Lukin's Harvard group ran protected operations on many logical qubits at once and fed in magic states. Another collaboration used three dimensional code structures to make magic states more cheaply. The authors note that every scheme still needs extra atoms for the checks

Reducing overhead is one of the most important goals in quantum computing right now. Neutral atoms are well suited because atoms can be moved around to connect distant qubits. This kind of work could shrink the size of useful machines considerably

Are neutral atoms the most promising route to fault tolerance? And how much can clever codes reduce hardware requirements?

Zhou's Group's MIT-Harvard Team Designs Compact Quantum Error Correction Codesquantumzeitgeist.com