Theory Seminar: Tensor-network methods for topological error correction

Andrew Darmawan
A universal quantum computer will require error correction to protect logical qubits from noise

Andrew Darmawan
Special theory group seminar, Herbert-Walther Lecture Hall G0.25
Mon 15. July 2019, 11:30 am

Abstract:

A universal quantum computer will require error correction to protect logical qubits from noise. However, practical implementation of quantum error correcting codes remains a challenge due to the large amount of physical resources (qubits, operations) required. I will discuss how some of these challenge may be met using tensor-network methods. I will focus on surface-code error correction, which, due to its simplicity, may be implementable in near-term devices. I will describe two ways in which tensor networks may be applied to surface-code error correction. One is for simulation: to understand how the surface code performs under different types of realistic noise. Another is as part of the classical control software of the error-correcting code, namely the decoder, which determines how to correct errors based on information provided by measurements. I will describe various scenarios in which tensor-network based decoders can achieve substantially improved logical error rates over other decoders.