Cost-effective Monitoring of Quantum Software Applications

量子计算机需要配备在其上运行的丰富而可靠的软件堆栈，才能充分发挥其解决实际问题的潜力。一个需要回答的根本问题是如何以经济高效的方式确保量子软件的正确和可靠。监控对于验证软件质量和诊断软件错误至关重要。然而，监控量子软件是独特的，也是具有挑战性的。例如，测量量子变量(即量子比特或量子比特)的状态将使它们坍塌成经典比特，并破坏它们的量子性质，如叠加。此外，监控量子应用程序需要额外的计算和存储资源(即，对量子比特进行操作的量子比特和量子门)。由于量子计算资源在可预见的未来仍然非常有限，确保监测不会带来太多的性能开销(例如，额外的量子比特和操作)是至关重要的。该项目旨在通过研究可靠且经济高效地监控量子软件应用程序的中间状态的方法来应对这些挑战。我们的第一个目标是设计可重复使用的量子监控组件，能够可靠地推断量子比特的状态，同时保持它们的保真度(即，最小化监控的功能影响)。我们的第二个目标是设计一个框架，以最优的方式将监控组件插入量子应用程序中，以最大限度地减少监控对性能的影响。该项目的成果将为提高量子软件应用程序的可观测性提供机会，这反过来将帮助加拿大的量子计算研究人员和量子软件开发人员了解其应用程序的运行时行为，并有效定位运行时错误。我们的项目将在确保量子软件应用程序的质量方面发挥重要作用，从而确保加拿大在量子计算方面的优势。

Quantum computers need to be equipped with a rich and reliable software stack running on top of them to reach their full potential in solving practical problems. A fundamental question that needs to be answered is how to cost-effectively ensure that quantum software is correct and reliable. Monitoring is essential to verify the quality of software and diagnose software errors. However, monitoring quantum software is unique and challenging. For example, measuring the states of quantum variables (i.e., quantum bits, or qubits) would collapse them into classical bits and destroy their quantum properties such as superposition. Additionally, monitoring a quantum application requires extra computing and storage resources (i.e., qubits and quantum gates that operate on the qubits). As quantum computing resources are and will continue to be very limited in the foreseeable future, it is critical to ensure that monitoring does not introduce too much performance overhead (e.g., extra qubits and operations). This project aims to address these challenges by investigating methods to reliably and cost-effectively monitor the intermediate states of quantum software applications. Our first goal is to design reusable quantum monitoring components that can reliably infer the states of qubits while preserving their fidelity (i.e., minimizing the functional impact of monitoring). Our second goal is to design a framework to optimally instrument the monitoring components into quantum applications to minimize the performance impact of monitoring. The outcomes of this project will provide opportunities to improve the observability of quantum software applications, which will in turn help quantum computing researchers and quantum software developers in Canada understand the runtime behaviors of their applications and effectively locate runtime errors. Our project will play an important part in assuring the quality of quantum software applications and hence, ensuring Canada's advantage in quantum computing.