Embedding quantum optimization problems using AC driven quantum ferromagnets

TL;DR

This paper introduces a novel AC-driven tunneling method called Symphonic Tunneling to mitigate embedding slowdowns in quantum optimization, potentially enhancing quantum speedups on near-term hardware.

Contribution

It proposes and demonstrates that AC variation of qubit parameters significantly accelerates tunneling, overcoming embedding-related slowdowns in quantum annealing.

Findings

AC-driven tunneling outperforms DC in speed

Symphonic Tunneling reduces embedding bottlenecks

Method extends to multi-chain clusters

Abstract

Analog quantum optimization methods, such as quantum annealing, are promising and at least partially noise tolerant ways to solve hard optimization and sampling problems with quantum hardware. However, they have thus far failed to demonstrate broadly applicable quantum speedups, and an important contributing factor to this is slowdowns from embedding, the process of mapping logical variables to long chains of physical qubits, enabling arbitrary connectivity on the short-ranged 2d hardware grid. Beyond the spatial overhead in qubit count, embedding can lead to severe time overhead, arising from processes where individual chains ``freeze" into ferromagnetic states at different times during evolution, and once frozen the tunneling rate of this single logical variable decays exponentially in chain length. We show that this effect can be substantially mitigated by local AC variation of the qubit parameters as in the RFQA protocol (Kapit and Oganesyan, Quant. Sci. Tech. \textbf{6}, 025013 (2021)), through a mechanism we call Symphonic Tunneling. We provide general arguments and substantial numerical evidence to show that AC-driven multi-qubit tunneling is dramatically faster than its DC counterpart, and since ST is not a 1d-specific mechanism, this enhancement should extend to clusters of coupled chains as well. And unlike a uniform transverse field, in higher dimensions this method cannot be efficiently simulated classically. We explore schemes to synchronize the AC tones within chains to further improve performance. Implemented at scale, these methods could significantly improve the prospects for achieving quantum scaling advantages in near-term hardware.