Taming Rydberg Decay with Measurement-based Quantum Computation

TL;DR

This paper introduces a measurement-based quantum computation approach to mitigate Rydberg decay errors in neutral atom quantum computers, enhancing error correction and reducing experimental complexity.

Contribution

It presents a novel scheme exploiting topological cluster states and leakage detection to address Rydberg decay errors without complex mid-circuit detection.

Findings

Achieves a 3.65% error threshold per CZ gate for Rydberg decay.

Demonstrates a high error distance comparable to the code distance.

Reduces experimental overhead compared to existing erasure conversion protocols.

Abstract

Programmable neutral atom arrays show great promise for fault-tolerant quantum computing. A dominant physical error on this platform is qubit leakage and loss, notably decay errors from the Rydberg state during two-qubit gates. Such leakage events are particularly detrimental as they propagate, generating correlated errors that severely degrade the effective error distance of quantum error correction codes. Here, we present a novel approach to address Rydberg decay errors leveraging measurement-based quantum computation (MBQC). Our scheme strategically exploits the inherent geometric structure of topological cluster states and only uses final leakage detection information to locate propagated errors originating from Rydberg decay. This eliminates the need for complex and atom-species-specific mid-circuit leakage detection, offering broader applicability, e.g., to the well-established Rb atom platform. We demonstrate a high error threshold of 3.65\% per CZ gate for pure Rydberg decay and achieve a favorable error distance $d_e \approx d$. Our method compares favorably with state-of-the-art erasure conversion protocols in the sub-threshold performance, offering comparable or marginally larger logical error rates while significantly reducing experimental overhead.