Suppressing quantum errors by scaling a surface code logical qubit

Quantum error correction offers a path to quantum computation at scale. By encoding a logical qubit into many physical qubits, we can protect against the unwanted effects of gate error and decoherence. Increasing the number of physical qubits increases this protection, but also increases the number of opportunities for errors to occur. This struggle between 'increased protection' and 'increased opportunity for error' is at the heart of large-scale fault-tolerant quantum computing.

In this talk, I'll present our most recent error-correction experiment on a 72-qubit Sycamore device. By scaling a 17-qubit distance-3 surface code to a 49-qubit distance-5 surface code, the 'increased protection' overcomes the 'increased opportunity for error' for the first time, with a modest ~4% average improvement in logical error rate. We'll discuss some of the improvements that made this experiment possible, and touch on some challenges to running a surface-code quantum computer at scale.

--- Michael Newman is a Senior Research Scientist at Google Quantum AI working on quantum error-correction and fault-tolerant quantum computing. He received his PhD in Mathematics from the University of Michigan and was previously a Postdoctoral Associate at Duke University.

---Co-hosted with IBM Quantum Hub at NC State University and Kenan Institute of Private Enterprise at UNC Kenan-Flagler Business School.