Hybrid compilation framework to accelerate quantum application development (CATALYST)

加速量子应用开发的混合编译框架（CATALYST）

• 批准号: 10071684
• 金额: $ 55.36万
• 依托单位: ROLLS-ROYCE PLC
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 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=10071684
• 关键词: Hybrid; compilation; framework; accelerate; quantum

Quantum computers are new types of powerful computers that are based on building blocks called qubits, that carry information in a more effective way than the bits on a conventional computer. Quantum computers have the potential to achieve computational times that are orders of magnitude faster than conventional computers.While qubits work differently from conventional bits, the computation workflow is somewhat similar: when we wish to run an algorithm, we write some lines of code that get translated into the so-called quantum circuit which then enacts a series of operations on the qubits before delivering a result. However, at the moment, this process is far from optimal, and the times needed for the pre-quantum steps required to run a calculation are prohibitively high. For quantum computers to become commercially useful, we need to not only optimise the algorithms we want to run and minimise the resources they need, but also reduce the time needed to translate them into a series of operations that can be then run on the qubits.While this constitutes an important problem for quantum computers, so far very little work has been done to address it. Another problem is that the quantum industry is currently very fragmented and still in its early stage of development.This project brings together leading quantum software and hardware companies from the UK and Canada - Riverlane and Xanadu - to help solve the technical challenge of improving the quality of the quantum algorithms and making them run easier, faster and better on the qubits. Riverlane will work on implementing techniques that allow algorithms to run using less resources, while Xanadu will develop a new hybrid classical-quantum compiler that will significantly decrease the calculation times and will allow users to use the appropriate resources in an optimal way. Rolls-Royce, a leader in power and propulsion systems will lead this project, providing real-world testcases that cannot be solved by today's quantum computers. Rolls-Royce will also develop new application software to best exploit the Riverlane and Xanadu developments.The partners will work together to combine improvements in quantum software, hardware and algorithms to significantly improve the runtime and results when running quantum algorithms. Our project brings together companies from UK and Canada to help develop quantum computers that will transform the way several sectors, such as finance, pharmaceuticals, aerospace etc. design and develop their products.

量子计算机是一种新型的强大计算机，基于称为量子位的构建块，它以比传统计算机上的位更有效的方式携带信息。量子计算机有可能实现比传统计算机快几个数量级的计算时间。虽然量子位的工作方式与传统位不同，但计算工作流程有些相似：当我们希望运行算法时，我们编写一些代码行，这些代码被翻译成所谓的量子电路，然后在提供结果之前对量子位执行一系列操作。然而，目前，这个过程远非最佳，并且运行计算所需的前量子步骤所需的时间非常长。为了让量子计算机变得商业化，我们不仅需要优化我们想要运行的算法并最大限度地减少它们所需的资源，而且还需要减少将它们转换为一系列可以在量子位上运行的操作所需的时间。虽然这对量子计算机来说是一个重要问题，但到目前为止，解决它的工作还很少。另一个问题是，量子产业目前非常分散，仍处于发展的早期阶段。该项目汇集了英国和加拿大领先的量子软硬件公司——Riverlane和Xanadu——帮助解决提高量子算法质量并使其在量子比特上运行更简单、更快、更好的技术挑战。 Riverlane 将致力于实现允许算法使用更少资源运行的技术，而 Xanadu 将开发一种新的混合经典量子编译器，它将显着减少计算时间，并允许用户以最佳方式使用适当的资源。动力和推进系统领域的领导者劳斯莱斯将领导这个项目，提供当今量子计算机无法解决的现实世界测试用例。罗尔斯·罗伊斯还将开发新的应用软件，以最好地利用 Riverlane 和 Xanadu 的开发成果。合作伙伴将共同努力，结合量子软件、硬件和算法的改进，以显着提高运行量子算法时的运行时间和结果。我们的项目汇集了来自英国和加拿大的公司，帮助开发量子计算机，这将改变金融、制药、航空航天等多个行业设计和开发产品的方式。