Expanding the benefits of quantum computing to new domains remains a challenging task. Quantum applications are concentrated in only a few domains, and driven by these few, the quantum stack is limited in supporting the development or execution demands of new applications. In this work, we address this problem by identifying both a new application domain, and new directions to shape the quantum stack. We introduce computational cognitive models as a new class of quantum applications. Such models have been crucial in understanding and replicating human intelligence, and our work connects them with quantum computing for the first time. Next, we analyze these applications to make the case for redesigning the quantum stack for programmability and better performance. Among the research opportunities we uncover, we study two simple ideas of quantum cloud scheduling using data from gate-based and annealing-based quantum computers. On the respective systems, these ideas can enable parallel execution, and improve throughput. Our work is a contribution towards realizing versatile quantum systems that can broaden the impact of quantum computing on science and society.

翻译：将量子计算的优势拓展至新领域仍是一项具有挑战性的任务。当前量子应用仅集中于少数领域，受此局限，量子计算栈难以支持新应用的开发或执行需求。针对这一问题，本文通过识别新应用领域及量子计算栈的优化方向展开研究。我们首次提出将计算认知模型作为新型量子应用类别——这类模型对人类智能的理解与复制至关重要，而本文首次将其与量子计算相联结。继而，我们通过分析这些应用，论证了通过重构量子计算栈以提升可编程性与性能的必要性。在发现的研究机遇中，我们基于门型与退火型量子计算机的数据，探索了两种量子云调度方案。在对应系统上，这些方案可实现并行执行并提升吞吐量。本工作致力于构建通用型量子系统，以拓展量子计算对科学与社会的深远影响。