意大利特伦托大学Philipp Hauke小组近日取得一项新成果。他们的最新研究提出了量子优化中真正的多部纠缠。相关论文于2025年2月20日发表在《物理评论A》杂志上。

据悉，在量子计算技术中产生二部纠缠的能力被广泛认为是关键。然而，真正的多部纠缠的作用比二部纠缠要少得多，特别是在解决量子器件复杂优化问题的背景下。从算法和硬件的角度来看，了解多方纠缠是否有助于，实现一个好的解决方案是至关重要的。

本文通过分析在Trotterized量子退火和量子近似优化算法中产生的，由广义几何测量量化的真多部纠缠来解决这一挑战。利用数值基准，研究人员详细分析了其在退火程序中的出现情况。研究组观察了多方纠缠障碍，并探讨了它与算法成功的关系。

研究组还证明了多部纠缠如何用精确解，提供了瞬时态重叠的上界。反之，与初始和最终产物状态的重叠，可以很容易地通过实验测量，为整个过程中的多部纠缠提供了上界。他们的结果有助于揭示，复杂的量子相关性如何成为量子优化的资源。

附：英文原文

Title: Genuine multipartite entanglement in quantum optimization

Author: Gopal Chandra Santra, Sudipto Singha Roy, Daniel J. Egger, Philipp Hauke

Issue&Volume: 2025/02/20

Abstract: The ability to generate bipartite entanglement in quantum computing technologies is widely regarded as pivotal. However, the role of genuinely multipartite entanglement is much less understood than bipartite entanglement, particularly in the context of solving complicated optimization problems using quantum devices. It is thus crucial from both the algorithmic and hardware standpoints to understand whether multipartite entanglement contributes to achieving a good solution. Here we tackle this challenge by analyzing genuine multipartite entanglement—quantified by the generalized geometric measure—generated in Trotterized quantum annealing and the quantum approximate optimization algorithm. Using numerical benchmarks, we analyze its occurrence in the annealing schedule in detail. We observe a multipartite-entanglement barrier, and we explore how it correlates to the algorithm's success. We also prove how multipartite entanglement provides an upper bound to the overlap of the instantaneous state with an exact solution. Vice versa, the overlaps to the initial and final product states, which can be easily measured experimentally, offer upper bounds for the multipartite entanglement during the entire schedule. Our results help to shed light on how complex quantum correlations come to bear as a resource in quantum optimization.

DOI: 10.1103/PhysRevA.111.022434

Source: https://journals.aps.org/pra/abstract/10.1103/PhysRevA.111.022434