Property Testing for Quantum Engineering (ProTeQE)

量子工程性能测试 (ProTeQE)

• 批准号: EP/X018180/1
• 负责人: Animesh Datta
• 金额: $ 25.79万
• 依托单位: University of Warwick
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=EP%2FX018180%2F1
• 关键词: Property; Testing; Quantum; Engineering; ProTeQE

Our senses incessantly inundate us with information on a multitude of issues in multifarious forms. Yet, when it comes to deciding any given topic, it is seldom feasible to use all of this information. Rather, we decide by deliberating on a selected, typically small, part of the whole.This self-evident realisation raises some fundamental questions. For instance, given a topic, is it even possible to make a meaningful decision by considering a small part of all the available information? Even if it were, how should this small part be selected?Theoretical computer science seeks to formalise these questions, and then answer them for specific mathematical problems. This forms the subject of property testing, which is concerned with the design of ultra-fast algorithms for approximate decision making. In this context, an approximate decision means choosing between objects that have a predetermined property and those that are 'far' from having that property. Applications range from checking the correctness of extremely large and complex software programmes to detecting pathologies based on interactions between proteins in the human body. For instance, deciding the correctness of a software by checking every line of code (which for a modern car can exceed 100 million lines) would be prohibitively expensive, time-consuming, and itself error prone. The ultra-fast property testing algorithms inspect only relatively small and random portions of these huge objects.ProTeQE will apply property testing to quantum engineering.Quantum computers are believed to be capable of solving several problems exponentially more efficiently than classical computers. Unsurprisingly, this comes at an exponential cost. Describing a quantum computer operating of n quantum bits (qubits) typically requires 4^n numbers. Once such a system is engineered, how can we decide if it can be used to build a reliable quantum computer?A reliable quantum computer - technically called a fault-tolerant quantum computer (FTQC) is only possible if the engineered quantum system obeys a specific set of mathematical properties. ProTeQE has the objective of answering the question: Given an object, in this case an engineered quantum system, does it have the properties necessary for FTQC, or is it far from having them?An ultra-fast algorithm to answer this question will remove a severe bottleneck in the development of quantum engineering and quantum technologies - that of testing quantum devices. The present testing regimen of measuring and processing exponentially many numbers to make a decision is prohibitively expensive, time-consuming, and itself error prone. ProTeQE will thus enable a faster turnaround in the designing and prototyping of engineered quantum systems. This should accelerate the development of reliable real-world quantum computers, and facilitate greater advances in quantum engineering and technologies more generally.ProTeQE should also nourish our basic curiosity. Quantum mechanics, which is presently our fundamental theory of Nature, is inherently probabilistic and non-local. When these concepts interface with those of property testing and approximate decision-making, the outcome could impact the foundations of our understanding of the laws of Nature. In particular, ProTeQE may eventually shed light on an abiding question: Are all fundamental laws of Nature (such as quantum mechanics) efficiently testable?

我们的感官不断地以各种形式向我们提供有关各种问题的信息。然而，当涉及到决定任何给定的主题时，使用所有这些信息很少可行。相反，我们通过对整体中一个选定的、通常很小的部分进行审议来做出决定，这一不言自明的认识提出了一些基本问题。例如，给定一个主题，是否有可能通过考虑所有可用信息的一小部分来做出有意义的决定？就算是，这一小部分又该如何挑选呢？理论计算机科学试图将这些问题形式化，然后用具体的数学问题来回答它们。这就形成了属性测试的主题，它与近似决策的超快速算法的设计有关。在这种情况下，近似决策意味着在具有预定属性的对象和远离该属性的对象之间进行选择。应用范围从检查极其庞大和复杂的软件程序的正确性到基于人体蛋白质之间的相互作用检测病理。例如，通过检查每一行代码（现代汽车可能超过1亿行）来确定软件的正确性将是非常昂贵，耗时且本身容易出错的。超快速属性测试算法仅检查这些巨大物体中相对较小和随机的部分。ProTeQE将属性测试应用于量子工程。量子计算机被认为能够比经典计算机以指数级更有效地解决几个问题。毫不奇怪，这是以指数级的成本来实现的。描述一台运行n个量子比特（qubit）的量子计算机通常需要4^n个数字。一旦这样的系统被设计出来，我们如何决定它是否可以用来建造可靠的量子计算机？可靠的量子计算机-技术上称为容错量子计算机（FTQC），只有在工程量子系统遵守一组特定的数学属性时才有可能。ProTeQE的目标是回答这样一个问题：给定一个物体，在这种情况下是一个工程量子系统，它是否具有FTQC所需的属性，或者它离这些属性还很远？回答这个问题的超快速算法将消除量子工程和量子技术发展中的一个严重瓶颈-测试量子设备。目前的测试方案，测量和处理指数级的许多数字，以作出决定是昂贵的，耗时的，本身容易出错。因此，ProTeQE将使工程量子系统的设计和原型设计能够更快地周转。这将加速可靠的现实世界量子计算机的发展，并促进量子工程和技术的更大进步。ProTeQE还将滋养我们基本的好奇心。量子力学是目前我们的基本自然理论，它本质上是概率性和非局部性的。当这些概念与属性测试和近似决策的概念相结合时，其结果可能会影响我们对自然规律的理解。特别是，ProTeQE最终可能会揭示一个永恒的问题：所有的自然基本定律（如量子力学）都是有效测试的吗？

项目成果

Provable Advantage in Quantum PAC Learning

量子 PAC 学习中可证明的优势

• DOI: 10.48550/arxiv.2309.10887
• 发表时间: 2023
• 作者: Salmon W

Accreditation of analogue quantum simulators.

模拟量子模拟器的认证。

• DOI: 10.1073/pnas.2309627121
• 发表时间: 2024
• 期刊: Proceedings of the National Academy of Sciences of the United States of America
• 影响因子: 11.1
• 作者: Jackson A