Establishing and exploiting Quantum advantage in superconducting Quantum annealer

新兴的量子计算领域面临的关键挑战之一是量子优势问题：通过直接利用量子力学定律，新设备是否能够比在经典硬件上运行的最佳算法更有效地执行某些计算任务？对于一个特定的量子设备来说，回答这个问题已经被证明是非常困难的，部分原因是量子优势通常是用人为的问题来测试的，这些问题设计得对量子设备来说很容易处理，但很难用经典的方法来解决。对于这样的问题，最好的经典算法可能是什么并不总是显而易见的，当发现和部署更好的经典算法时，关于量子优势的结论可以在以后被逆转。在拟议的研究计划中，我们将通过采用最先进的张量网络模拟技术，研究D-Wave系统（一家位于bc省本那比的量子计算公司）开发的量子退火器中的量子优势。退火架构实现了横向场Ising模型，在超导量子比特和非常灵活的连接之间具有单独可调的耦合，能够模拟具有高达~5000个量子自旋的1、2和3维系统。与某些经典算法相比，D-Wave团队已经建立了量子优势的各个方面，包括精确数值对角化和量子蒙特卡罗。我们基于ubc的团队在张量网络（TN）方法方面拥有丰富的经验，尽管它是已知最强大的算法，但尚未应用于此问题。在与D-Wave团队成员的密切合作下，我们将努力调整和完善现有的TN方法，以完成量子退退火的基准测试任务，然后分析两者之间直接比较的结果，最后通过解决量子凝聚态物理和实际应用中的难题来利用优势。预期的结果将推动加拿大的量子退火努力在这一竞争领域处于世界领先地位。

One of the key challenges facing the nascent field of quantum computation is the question of quantum advantage: By directly exploiting the laws of quantum mechanics is the new device capable of performing certain computational tasks more efficiently than the best available algorithm running on classical hardware? Answering this question for a particular quantum device has proven surprisingly difficult, in part because quantum advantage is often tested using artificial problems designed to be tractable for the quantum device but difficult to solve classically. For such problems it is not always obvious what the best classical algorithm might be and the verdict regarding quantum advantage can be reversed later when a better classical algorithm is found and deployed. In the proposed program of research we will study the quantum advantage in the quantum annealer developed by D-Wave Systems, a Burnaby BC-based quantum computing company, by employing state-of-the-art Tensor Network simulation techniques. The annealer architecture implements the transverse-field Ising model with individually adjustable couplings between superconducting qubits and very flexible connectivity, capable of simulating 1, 2 and 3-dimensional systems with up to ~5000 quantum spins. The D-Wave team has already established various aspects of quantum advantage as compared to certain classical algorithms, including the exact numerical diagonalization and quantum Monte Carlo. Our UBC-based team possesses extensive experience in Tensor Network (TN) methods which have not yet been applied to this problem despite being the most powerful known algorithm. In close collaboration with D-Wave team members we will work to adapt and refine the existing TN methods to the task of benchmarking the quantum annealer, then analyze the outcome of direct comparisons between the two and finally exploit the advantage by solving hard problems in quantum condensed matter physics and in real-life applications. The expected outcome will propel the Canadian quantum annealing effort to the world-leading position in this competitive field.