State-dependent Routing Dynamics in Noisy Quantum Computing Devices

TL;DR

This paper introduces a model for understanding how routing decisions in noisy quantum computers affect state fidelity, enabling more effective routing strategies in NISQ devices.

Contribution

It develops a state-dependent noise model for quantum routing operations that can be efficiently characterized and used to optimize routing in NISQ devices.

Findings

The model accurately predicts state-dependent noise dynamics.

Pair-wise measurements suffice for efficient model characterization.

Model predictions align with experimental state reconstructions.

Abstract

Routing plays an important role in programming noisy, intermediate-scale quantum (NISQ) devices, where limited connectivity in the register is overcome by swapping quantum information between locations. However, routing a quantum state using noisy gates introduces non-trivial noise dynamics, and deciding on an optimal route to minimize accumulated error requires estimates of the expected state fidelity. Here we validate a model for state-dependent routing dynamics in a NISQ processor based on correlated binary noise. We develop a composable, state-dependent noise model for CNOT and SWAP operations that can be characterized efficiently using pair-wise experimental measurements, and we compare model predictions with tomographic state reconstructions recovered from a quantum device. These results capture the state-dependent routing dynamics that are needed to guide routing decisions for near-real time operation of NISQ devices.