Multicore NISQ Processors on Silicon Chips

量子计算机代表着在最深层次上利用自然，根据我们今天所知的物理定律，建造我们能想象到的最强大的计算机器。据预测，他们将改变从物流到材料和药物的发现以及安全等领域。量子计算最深远的影响将需要完全纠正计算中的错误，而这种能力预计需要多达数百万个量子比特，或“量子位”，所有这些都是通过量子链路连接起来的。然而，越来越多的证据表明，即使是相对小规模的量子处理器，如果没有纠错，也能够解决有用的问题，并提供颠覆性的进步。例如，一台只有53个基本量子比特（没有纠错）的量子计算机最近在解决计算问题的竞赛中击败了世界上最强大的超级计算机。然而，所选择的计算问题是一个没有实际价值的人为问题，旨在支持量子计算机，并且使用这种小规模量子处理器来解决有用的问题并实现一些人所说的“量子优势”仍然是一个开放和重要的挑战。增强小型量子处理器能力的一种方法是并行操作它们——本质上是使用量子处理器的许多副本，并赋予它们相关的任务来解决。在实践中，这种“多核”方法可以为材料和药物建模的量子算法设计提供实质性的加速。然而，这种方法隐含着较低的“每量子位成本”，这使得制造许多独立的量子处理器成为可能，并且能够将量子处理器连接到传统计算机进行控制。硅为量子计算提供了一个非常适合这种方法的平台，能够利用CMOS技术生产量子位，以及将量子处理器连接到所需控制器的传统电子设备。在这个可行性项目中，我们将进一步发展硅中的多核量子处理器概念，从实验和理论上，确定如何使用CMOS技术实现它，以及它的预测能力将是什么，以及它将能够解决什么新问题。

Quantum computers represent harnessing nature at its deepest level to build the most capable computing machines we can imagine based on the laws of physics we know today. They have been predicted to transform areas ranging from logistics, to the discovery of materials and drugs, and security. The most profound impacts of quantum computing will require the full correction of errors in the calculation, and this capability is expected to require up to millions of quantum bits, or 'qubits', all connected by quantum links. However, there is mounting evidence that even relatively small-scale quantum processors, without error correction, will be capable of solving useful problems and offering disruptive advances. For example, a quantum computer with just 53 elementary quantum bits (and no error correction) has recently beaten the world's most powerful supercomputer in a competition to solve a computation problem. However, the computation problem chosen was a contrived one of no practical value, designed to favour the quantum computer, and it remains an open and important challenge to use such small-scale quantum processors to solve useful problems and achieve what some have termed 'quantum advantage'.One way to enhance the power of small-scale quantum processors is to operate them in parallel -- essentially taking many copies of the quantum processor and giving them related tasks to solve. In practice, this 'multi-core' approach can offer substantial speed-ups for quantum algorithms design for modelling materials and drugs. However, implicit in this approach is a low 'cost per qubit', which allows the manufacture of many independent quantum processors, and the ability to interface the quantum processors to a conventional computer for control. Silicon offers a platform for quantum computing which is ideally suited to this approach, being able to leverage CMOS technology to produce qubits, as well as the conventional electronics to connect the quantum processors to the required controller. In this feasibility project, we will further develop the multi-core quantum processor concept in silicon, both experimentally and theoretically, to establish how it can be realised using CMOS technology and what its predicted capabilities will be and what new problems it will be able to solve.