Noise-avoidance and Simulation in Quantum Information Technologies

量子技术提供的进步将从根本上影响21世纪。在过去的两年里，随着量子计算进入了一个称为嘈杂中尺度量子(NISQ)时代的阶段，出现了一场“量子空间竞赛”，来自主要工业方向(如Google和IBM)和世界各国(美国：约)的兴趣激增。每年2亿美元&12亿美元的量子技术投资。中国：合肥实验室10亿美元)。工作进展迅速，10月份，谷歌宣布了一个关键的里程碑：第一个在量子设备上进行的计算，大大超过了世界上最大的超级计算机的能力。这些进步不仅仅是提高了硬件效率，相反，量子计算机对应着对计算含义的根本重新思考。一种说法是，量子计算机对于超级计算机就像超级计算机对于算盘一样，因此它的变革潜力可能深刻地塑造我们的世界。量子技术的广泛应用可以解决我们世界上一些最大的问题。一个突出的例子是哈伯-博世化肥生产工艺，其效率仅为60%，约占全球能源消耗的3%。然而，大自然对于同样的任务几乎实现了100%的效率。我们不能模仿大自然的方法的一个核心原因是，所使用的分子超出了超级计算机的模拟能力。然而，这正是量子计算机能够以指数级速度模拟的那种强相互作用的复杂系统。这只是一个潜在的应用，但还有许多其他的应用，如目前难以处理的化学模拟，医学上的新药发现，或新超导材料的开发。现在的一个根本挑战是如何为NISQ时代设计量子器件，并在面临噪声和硬件限制(例如，设备组件的有限连接)的情况下以最佳方式利用这种量子系统。这一理论研究计划将为中期量子设备开发新的协议，这些设备在嘈杂的环境中尽可能优化地利用宝贵的量子资源，并能够自然地纳入特定的硬件限制。核心方法论采用了最近为量子系统热力学开发的方法，并以新颖的方式将它们应用于量子技术目标。它将首先通过使用谐波分析来对量子设备的性能进行基准测试。这一点很重要，因为现有的量子计算机噪音很大，准确的诊断是必不可少的。其次，它将使用最新的统计力学工具来量化对全面量子计算至关重要的脆弱的量子组件(称为“魔力状态”)。这一点很重要，因为目前对魔态的要求非常高，因此这一领域的任何进展都将极大地加快向全尺寸量子计算机的进展。该计划还将使用热力学第二定律的最新变体来开发量子纠错公式(健壮量子计算机的另一个关键组件)，这些公式明确地为实验测试量身定做。最后，它将利用在该计划早期阶段获得的见解来开发新的经典算法(用于普通、非量子计算机)。这些经典算法将有助于我们模拟量子系统(例如，估计量子计算机变得有用的计算机制)，并与现有的药物发现算法工作相联系。它将通过利用量子统计力学的最新扩展来做到这一点。该方法非常适合，因为它提供了一种与传统统计力学非常成功地应用于经典计算(例如蒙特卡罗方法)的现代平行，同时也涉及到硬件挑战。

Quantum Technologies offer advances that will impact the 21st Century in fundamental ways. The past 2 years have seen the emergence of a "quantum space race" with quantum computing entering a phase called the Noisy Intermediate-Scale Quantum (NISQ) era, with an explosion of interest from prominent industrial directions (e.g. Google & IBM) and countries around the world (USA: approx. $200 million/year & $1.2bn investment for quantum technologies. China: $1 billion for the Hefei labs). Work is progressing quickly, and in October a key milestone was claimed by Google: the first computation on a quantum device that dramatically outperformed the capabilities of the world's largest supercomputer. These advances are more than just improved hardware efficiency, instead a quantum computer corresponds to a fundamental re-thinking of what computing means. A claim is that a quantum computer will be to a supercomputer what a supercomputer is to an abacus, and as such its transformative potential could profoundly shape our world. There is a broad spectrum of applications of quantum technologies that could address some of our world's biggest problems. A prominent example is the Haber-Bosch process for fertiliser production which has an efficiency of only 60%, and responsible for about 3% of global energy consumption. Nature, however, achieves close to 100% efficiency for the same task. A core reason we cannot mimic Nature's method is that the molecule employed is beyond the simulation abilities of supercomputers. However this is precisely the kind of strongly interacting complex system that quantum computers can simulate exponentially faster. This is just one potential application but there are many others such as currently intractable chemistry simulations, novel drug-discovery in medicine, or the development of new superconducting materials.A fundamental challenge now is how to design quantum devices for the NISQ era and optimally make use of such quantum systems in the face of noise & hardware limitations (e.g. limited connectivity of device components). This theory research programme will develop novel protocols for medium term quantum devices that exploit precious quantum resources in noisy environments as optimally as possible with the ability to naturally incorporate specific hardware limitations. The core methodology takes recent methods developed for the thermodynamics of quantum systems and applies them in novel ways to quantum technology goals. It will do this firstly by using harmonic analysis for the performance benchmarking of quantum devices. This is important as existing quantum computers are very noisy and a sharp diagnosis is essential. Secondly, it will use recent statistical mechanics tools to quantify fragile quantum components (called 'magic states') that are vital for full-scale quantum computing. This is important because current magic state demands are extremely high and so any advances in this line would greatly speed up progress to a full-scale quantum computer. The programme will also use a very recent variant of the 2nd Law of Thermodynamics to develop formulations of quantum error-correction (another key component of a robust quantum computer) that are explicitly tailored to experimental testing. Finally, it will use the insights gained in the early phase of the programme to develop new classical algorithms (for ordinary, non-quantum computers). These classical algorithms will contribute to our ability to simulate quantum systems (e.g. to estimate the computational regime where quantum computers become useful) and also connect with existing work on drug-discovery algorithms. It will do so by exploiting recent extensions of quantum statistical mechanics. The methodology is well-suited as it provides a modern parallel to the highly successful application of traditional statistical mechanics to classical computing (e.g. Monte Carlo methods) while also engaging hardware challenges.