This tutorial offers a practical introduction to the fascinating world of quantum computation and its application to optimization and machine learning problems. The participants will acquire hands-on experience in developing hybrid "Variational Quantum Algorithms", which combine classical and quantum computation, and in running them both on simulators and real quantum hardware provided by leading ICT companies. As a concrete use-case, of specific interest for the HPDC community, the tutorial will discuss the optimal assignment and scheduling of resources on the different nodes and layers of a Cloud/Edge architecture, a problem that is known to have NP-hard complexity.
This tutorial offers a practical introduction to the fascinating world of quantum computation and its application to optimization and machine learning problems. The participants will acquire hands-on experience in developing hybrid "Variational Quantum Algorithms", which combine classical and quantum computation, and in running them both on simulators and real quantum hardware provided by leading ICT companies. As a concrete use-case, of specific interest for the HPDC community, the tutorial will discuss the optimal assignment and scheduling of resources on the different nodes and layers of a Cloud/Edge architecture, a problem that is known to have NP-hard complexity.
Modern Cloud/Edge architectures need to orchestrate multiple layers of heterogeneous computing nodes, including pervasive sensors/actuators, distributed Edge/Fog nodes, centralized data centers and quantum devices. The optimal assignment and scheduling of computation on the different nodes is a very difficult problem, with NP-hard complexity. In this paper, we explore the possibility of solving this problem with Variational Quantum Algorithms, which can become a viable alternative to classical algorithms in the near future. In particular, we compare the performances, in terms of success probability, of two algorithms, i.e., Quantum Approximate Optimization Algorithm (QAOA) and Variational Quantum Eigensolver (VQE). The simulation experiments, performed for a set of simple problems, show that the VQE algorithm ensures better performances when it is equipped with appropriate circuitansatzesthat are able to restrict the search space. Moreover, experiments executed on real quantum hardware show that the execution time, when increasing the size of the problem, grows much more slowly than the trend obtained with classical computation, which is known to be exponential.
Abstract This practical guide presents hands-on laboratory exercises on variational quantum algorithms (VQAs) and hybrid quantum-classical workflows, based on the IBM Q learning platform. The work was performed by the author on April 28, 2024, using IBM Q cloud resources. Variational algorithms leverage the variational principle of quantum mechanics to optimize quantum circuits for specific tasks, making them particularly suitable for current noisy intermediate-scale quantum (NISQ) devices and offering a practical route toward achieving quantum advantage. The guide explores the full workflow for designing and implementing variational algorithms, highlighting the trade-offs inherent at each stage, including circuit design, parameter initialization, optimization strategies, and measurement post-processing. It also demonstrates the use of Qiskit Runtime primitives to enhance both the speed and accuracy of computations. Through these exercises, participants gain experience in constructing hybrid quantum-classical algorithms, performing parameter optimization, and interpreting results within the context of near-term quantum devices. While this guide focuses on practical implementation, it also encourages exploration of the underlying theoretical foundations of quantum information and computation, connecting applied algorithmic practice with fundamental principles. The course and exercises provide a structured starting point for researchers, developers, and students aiming to explore the utility of quantum computers for real-world problem solving, particularly in optimization, quantum chemistry, and machine learning applications.
Optimizing for a classification problem with a 0-1 loss function is an NP-hard problem [1] even for a simple binary classification task. As Preskill stated “We don't expect a quantum computer to solve worst case instances of NP-hard problems, but it might find better approximate solution or find it faster” [2]. Hybrid Quantum Classical Platforms harness the full capacity of Near-Term Quantum Computers. In this study, a focus is given on the variational circuits based approach accomplishing various machine learning tasks. We will survey a variety of experimental demonstrations conducted on actual quantum hardware and actively developing simulation software, anticipated to have a broad range of real-world applications in this increasingly growing field.
The convergence of artificial intelligence (AI), quantum computing, and cloud infrastructure represents a paradigm shift in how intelligent systems are designed, trained, and deployed. As AI models continue to demand exponential computational resources, classical systems are increasingly reaching their performance limits in domains such as high-dimensional optimization, complex pattern recognition, and probabilistic modeling. Quantum computing, with its foundational principles of superposition, entanglement, and quantum parallelism, offers a theoretical edge in solving problems that are intractable for conventional hardware. However, the practical deployment of quantum algorithms—particularly those relevant to machine learning (ML) has been hindered by the limited scalability and accessibility of quantum hardware. To bridge this gap, cloud-based quantum computing platforms have emerged, providing remote, programmable access to quantum devices and hybrid runtimes. In this context, the present study offers a comprehensive review of cloud-delivered quantum AI systems, examining core architectures, hybrid quantum-classical algorithms, platform capabilities, and real-world use cases. The paper synthesizes recent advances and performance insights from systems such as Amazon Braket, IBM Quantum, and Azure Quantum, while critically analyzing algorithmic trends like variational quantum circuits and quantum kernel methods. It further explores challenges including circuit depth limitations, latency in hybrid models, and standardization barriers across toolchains. By consolidating current research and implementation patterns, this review aims to inform the design and future trajectory of scalable, cloud-native quantum AI applications.
The chapter Hybrid Algorithms for Quantum Computing and Artificial Intelligence delves into the groundbreaking potential of hybrid algorithms that synergize quantum and classical computing to solve complex problems across various domains. It introduces foundational concepts, focuses on the interplay between quantum mechanics and classical computation, and explores task distribution strategies and quantum-classical computation models. Highlighting the advantages of hybrid systems over standalone approaches, the chapter examines advancements such as variational quantum algorithms, quantum-assisted optimization, and hybrid neural networks. These innovations are contextualized with practical applications in supply chain management, energy distribution, and machine learning, showcasing their impact on efficiency, accuracy, and scalability. Real-world case studies illustrate the transformative capabilities of these algorithms, addressing challenges like noise and error correction in quantum systems. This chapter provides a comprehensive overview of hybrid architectures, emphasizing their role as a bridge to the future of computational science and innovation.
Quantum computing is a rapidly developing field that has seen a huge amount of interest in the last couple of decades due to its promise of revolutionizing several domains of business and science. It presents a new way of doing computations by making use of fundamental properties of quantum mechanics such as superposition and entanglement. Optimization, on the other hand, is a field that is omnipresent in the industry and where small improvements can have a significant impact. This thesis aims to tackle optimization problems using quantum algorithms.NP-hard optimization problems are not believed to be exactly solvable through general polynomial time algorithms. Variational quantum algorithms (VQAs) to address such combinatorial problems have been of great interest recently. Such algorithms are heuristic and aim to obtain an approximate solution. The hardware, however, is still in its infancy and the current Noisy Intermediate Scale Quantum (NISQ) computers are not able to optimize industrially relevant problems. Moreover, the storage of qubits and introduction of entanglement require extreme physical conditions.An issue with contemporary quantum optimization algorithms such as the Quantum Approximate Optimization Algorithm (QAOA) is that they scale linearly with problem size. To tackle this issue, we present the LogQ encoding, using which we can design quantum variational algorithms that scale logarithmically with problem size - opening an avenue for treating optimization problems of unprecedented scale on gate-based quantum computers. We show how this algorithm can be applied to several combinatorial optimization problems such as Maximum Cut, Minimum Partition, Maximum Clique and Maximum Weighted Independent Set (MWIS). Subsequently, these algorithms are tested on a quantum simulator with graph sizes of over a hundred nodes and on a real quantum computer up to graph sizes of 256. To our knowledge, these constitute the largest realistic combinatorial optimization problems ever run on a NISQ device, overcoming previous problem sizes by almost tenfold.Next, we apply the LogQ encoding to two use-cases for large companies such as TotalEnergies. Fleet conversion is the process of transitioning a fleet of vehicles to more sustainable and environmentally friendly alternatives. It is modeled as a column generation scheme with the MWIS problem as the sub-problem or worker problem. We use the LogQ method to solve the MWIS Workers and demonstrate how quantum and classical solvers can be used together in a hybrid manner to approach an industrial-sized use-case. Mesh segmentation refers to the process of dividing a complex mesh (composed of vertices, edges, and faces) into meaningful and semantically coherent parts or regions. Mesh segmentation plays an important part in computer modeling, which is extensively used in the core domains of TotalEnergies' activities such as Earth imaging, physical modeling for reservoirs, and others. We define the problem as a graph optimization problem and use the LogQ encoding to solve it.