Quantum Computing (QC) has emerged as a field of ever-increasing activity as it promises to revolutionize computation and enable the solution of computational problems that we (realistically) cannot solve with Classical Computing to date. However, existing quantum programming environments mostly require an in-depth understanding of the basic QC building blocks, that is, quantum states, superposition, entanglement and measurement as well as the changing of quantum states using basic quantum gates. The present state of quantum programming reminds us of how Classical Computing was about 70 years ago when computing machines such as the ENIAC required significant effort to program solely using very basic digital building blocks. Over the decades, though, increasingly higher-level abstractions have been created on top of the basic building blocks of Classical Computing and made computation much more accessible and wide-spread. In order to make Quantum Computing more accessible, we argue that Software Engineering for QC needs to embark on a similar journey and create abstractions that shield developers from the basic QC building blocks as much as possible so that they can focus their attention on solving problems and less on how to manipulate quantum states using quantum circuits. Based on our experience of developing a scalable quantum n-queens solver, this paper aims to formulate recommendations for raising the level of abstraction in Quantum Software Engineering.
Abstraction layers are of paramount importance in software architecture, as they shield the higher-level formulation of payload computations from lower-level details. Since quantum computing (QC) introduces many such details that are often unaccustomed to computer scientists, an obvious desideratum is to devise appropriate abstraction layers for QC. For discrete optimisation, one such abstraction is to cast problems in quadratic unconstrained binary optimisation (QUBO) form, which is amenable to a variety of quantum approaches. However, different mathematically equivalent forms can lead to different behaviour on quantum hardware, ranging from ease of mapping onto qubits to performance scalability. In this work, we show how using higher-order problem formulations (that provide better expressivity in modelling optimisation tasks than plain QUBO formulations) and their automatic transformation into QUBO form can be used to leverage such differences to prioritise between different desired non-functional properties for quantum optimisation. Based on a practically relevant use-case and a graph-theoretic analysis, we evaluate how different transformation approaches influence widely used quantum performance metrics (circuit depth, gates count, gate distribution, qubit scaling), and also consider the classical computational efforts required to perform the transformations, as they influence possibilities for achieving future quantum advantage. Furthermore, we establish more general properties and invariants of the transformation methods. Our quantitative study shows that the approach allows us to satisfy different trade-offs, and suggests various possibilities for the future construction of general-purpose abstractions and automatic generation of useful quantum circuits from high-level problem descriptions.
Quantum computing has become an active research field in recent years, as its applications in fields such as cryptography, optimization, and materials science are promising. Along with these developments, challenges and opportunities exist in the field of Quantum Software Engineering, as the development of frameworks and higher-level abstractions has attracted practitioners from diverse backgrounds. Unlike initial quantum frameworks based on the circuit model, recent frameworks and libraries leverage higher-level abstractions for creating quantum programs. This paper presents an empirical study of 985 Jupyter Notebooks from 80 open-source projects to investigate how quantum patterns are applied in practice. Our work involved two main stages. First, we built a knowledge base from three quantum computing frameworks (Qiskit, PennyLane, and Classiq). This process led us to identify and document 9 new patterns that refine and extend the existing quantum computing pattern catalog. Second, we developed a reusable semantic search tool to automatically detect these patterns across our large-scale dataset, providing a practitioner-focused analysis. Our results show that developers use patterns in three levels: from foundational circuit utilities, to common algorithmic primitives (e.g., Amplitude Amplification), up to domain-specific applications for finance and optimization. This indicates a maturing field where developers are increasingly using high-level building blocks to solve real-world problems.
Existing abstract models of quantum computation make reference to circuit elements, much in contrast to their classical counterparts. Circuits, as a model of computation, substantially limit algorithmic expression and obscure high-level connections between problems and quantum resources. It is argued here that new models are needed to achieve high-level algorithmic expressiveness that allow composable procedural abstractions to manifest, leading to the development of instructions in the sense usually understood in high-level programming languages. Doing so appears essential to the discovery of new quantum algorithms, and deeper understanding of how quantum resources compose into useful patterns, or \emph{quantum motifs}. To achieve this, stronger investment in the intersection between higher-algebra, mathematical physics and quantum science is required to cope with future challenges brought forth by \textit{very large quantum scale integration}.