MACON-QC: Many-Body Phases In Continuous-Time Quantum Computation

量子计算的连续时间方法（QC）在系统堆栈的所有级别（特别是在软件级别）都与基于门的QC有根本的不同。一般来说，连续时间方法在退相干方面具有更强的鲁棒性，尽管目前的纠错技术落后于基于门的QC。然而，多量子位连续时间工程量子系统与本质上的凝聚态系统之间的明显类比允许软件设计师开发算法，这些算法受到上个世纪积累的凝聚态物理中量子相变的大量知识的启发和借鉴。在MACON-QC中，我们将开发连续时间算法，该算法直接受到凝聚态物质中多体相变的启发。这些算法将利用目前无法在硬件中实现的哈密顿量，即使是在两个量子位水平上。该研究的最终目标是确定需要哪些定性的新硬件元素，以便使未来的压缩物质启发的连续时间量子计算机能够在实际应用中展示量子加速。与此任务相结合，将开发在连续进化过程中证明真正量子行为的方法，以便可以证明算法经过真正的量子状态。还将探索通过机器学习优化新添加的元素和算法。

Continuous-time approaches to quantum computation (QC) are fundamentally different from gate-based QC at all levels of the system stack, particularly at the software level. In general continuous-time approaches can be more robust with respect to decoherence, though at present the state of art of error-correction lags behind that of gate-based QC. Nevertheless the clear analogies between many-qubit continuous-time engineered quantum systems and condensed-matter systems in nature allow software designers to develop algorithms which are inspired by and draw on the vast knowledge of quantum phase-transitions in condensed-matter physics accumulated over the last century.In MACON-QC we will develop continuous-time algorithms which are directly inspired by many-body phase transitions in condensed matter. These algorithms will utilise Hamiltonians which cannot currently be implemented in hardware, even at the two-qubit level. The ultimate goal of the research is to establish which qualitatively new hardware elements are required so as to enable future condensed-matter-inspired continuous-time quantum computers to demonstrate quantum speedup in real-world applications. In tandem with this task, approaches to certifying truly quantum behaviour during the continuous evolution will be developed so that it can be demonstrated that the algorithm goes through truly quantum states. Optimization of the newly-added elements and the algorithms via machine learning will also be explored.