The new quantum era is expected to have an unprecedented social impact, enabling the research of tomorrow in several pivotal fields. These perspectives require a physical system able to encode, process and store for a sufficiently long amount of time the quantum information. However, the optimal engineering of currently available quantum computers, which are small and flawed by several non-ideal phenomena, requires an efficacious methodology for exploring the design space. Hence, there is an unmet need for the development of reliable hardware-aware simulation infrastructures able to efficiently emulate the behaviour of quantum hardware that commits to looking for innovative systematic ways, with a bottom-up approach starting from the physical level, moving to the device level and up to the system level. This article discusses the development of a classical simulation infrastructure for semiconductor quantum-dot quantum computation based on compact models, where each device is described in terms of the main physical parameters affecting its performance in a sufficiently easy way from a computational point of view for providing accurate results without involving sophisticated physical simulators, thus reducing the requirements on CPU and memory. The effectiveness of the involved approximations is tested on a benchmark of quantum circuits — in the expected operating ranges of quantum hardware — by comparing the corresponding outcomes with those obtained via numeric integration of the Schrödinger equation. The achieved results give evidence that this work is a step forward towards the definition of a classical simulator of quantum computers.
The new quantum era is expected to have an unprecedented social impact, enabling the research of tomorrow in several pivotal fields. These perspectives require a physical system able to encode, process and store for a sufficiently long amount of time the quantum information. However, the optimal engineering of currently available quantum computers, which are small and flawed by several non-ideal phenomena, requires an efficacious methodology for exploring the design space. Hence, there is an unmet need for the development of reliable hardware-aware simulation infrastructures able to efficiently emulate the behaviour of quantum hardware that commits to looking for innovative systematic ways, with a bottom-up approach starting from the physical level, moving to the device level and up to the system level. This article discusses the development of a classical simulation infrastructure for semiconductor quantum-dot quantum computation based on compact models, where each device is described in terms of the main physical parameters affecting its performance in a sufficiently easy way from a computational point of view for providing accurate results without involving sophisticated physical simulators, thus reducing the requirements on CPU and memory. The effectiveness of the involved approximations is tested on a benchmark of quantum circuits — in the expected operating ranges of quantum hardware — by comparing the corresponding outcomes with those obtained via numeric integration of the Schrödinger equation. The achieved results give evidence that this work is a step forward towards the definition of a classical simulator of quantum computers.
Classical simulation of Noisy Intermediate Scale Quantum computers is a crucial task for testing the expected performance of real hardware. The standard approach, based on solving Schrödinger and Lindblad equations, is demanding when scaling the number of qubits in terms of both execution time and memory. In this article, attempts in defining compact models for the simulation of quantum hardware are proposed, ensuring results close to those obtained with standard formalism. Molecular Nuclear Magnetic Resonance quantum hardware is the target technology, where three non-ideality phenomena—common to other quantum technologies—are taken into account: decoherence, off-resonance qubit evolution, and undesired qubit-qubit residual interaction. A model for each non-ideality phenomenon is embedded into a MATLAB simulation infrastructure of noisy quantum computers. The accuracy of the models is tested on a benchmark of quantum circuits, in the expected operating ranges of quantum hardware. The corresponding outcomes are compared with those obtained via numeric integration of the Schrödinger equation and the Qiskit’s QASMSimulator. The achieved results give evidence that this work is a step forward towards the definition of compact models able to provide fast results close to those obtained with the traditional physical simulation strategies, thus paving the way for their integration into a classical simulator of quantum computers.
Quantum computing could impact Society and Industry in an unprecedented way, especially in critical domains, such as optimization problem solving, computational chemistry and artificial intelligence.Properties of quantum mechanics, such as superposition and entanglement, can be exploited to define problem-specific quantum algorithms with lower computational complexity than their best corresponding classical counterparts.A common limitation of quantum computers is the presence of hardware non-idealities, such as decoherence, dramatically affecting performance and reliability.In any case, efforts in improving hardware and investigating short and long-term quantum algorithms can be already done.The goal of this doctorate thesis is to explore different levels of quantum computing engineering, from devices to applications.In particular, the attention was focused on two activities: the definition of a compact-model-based simulation methodology for quantum technologies for the general-purpose quantum circuit model and the proposal of an approach for solving the users scheduling optimization problem in Joint-Transmission-based wireless communication systems, exploiting the quantumcompliant Quadratic Unconstrained Binary Optimization (QUBO) formalism.Classical simulation of quantum computers allows a preliminary estimation of their capabilities in specific applications.Canonical simulation of qubits affected by non-ideality phenomena requires solving the computationally expensive Lindblad master equation.A compact model of a quantum technology, obtained through fitting operations or specific mimic algorithms, is thought to describe qubits evolution under quantum gates, taking into account the device-characteristic physical parameters and not numerically solving Lindblad equation, thus permitting faster and more scalable simulations.Compact models of two families of spin-based molecular quantum computers, switch-based and Nuclear Magnetic Resonance, were developed, both involving two other general non-ideality models: decoherence and off-resonance of untargeted qubits.A quantum circuits simulator, involving all these compact models and interfaceable with OpenQASM 2.0 language, was entirely developed in MATLAB.The compact model approximations were compared with either results already available in the state of art or through numerical simulations with consolidated formalisms, at both Hamiltonian and circuit levels.The achieved results give evidence that the proposed simulation methodology provides iii results fast and close to the references, thus paving the way for fast and reliable technology-dependent quantum circuits simulation.Joint Transmission (JT) is a modern mechanism for Multiple-Input-Multiple-Output (MIMO) wireless communication systems, where multiple antennas co-operate to transmit signals to users with minimized interference.An optimization problem, looking for the best users subset servable in JT, can be defined.In a practical scenario, this cannot be efficiently solved with a brute-force-only approach, since it could imply a too high number of combinations.This thesis presents a strategy for solving the JT optimization problem, based on a QUBO model, which is responsible of a first selection of users among all available.The subset selected with QUBO is then brute-force explored, to find the best solution.In all executed tests, the obtained results permit to say that the proposed method achieves a quite good compromise between solution quality and total execution time.Even though the obtained results in both contexts can still be further expanded and improved, they are quite encouraging and prove that these research activities, if continued in the future, can help in improving quantum computing engineering, at both hardware and software levels, with the hope that it can concretely solve problems of interest for Society and Industry. Part I Introductory chapters 1 BackgroundIn this first section, an intuitive description of the main actors of quantum computing is provieded, with a historical timeline with the milestones of this computational paradigm.
Quantum computing promises to revolutionize several scientific and technological domains through fundamentally new ways of processing information. Among its most compelling applications is digital quantum simulation, where quantum computers are used to replicate the behavior of other quantum systems. This could enable the study of problems that are otherwise intractable on classical computers, transforming fields such as quantum chemistry, condensed matter physics, and materials science. Despite this potential, realizations of practical quantum advantage for relevant problems are hindered by imperfections of current devices. This also affects quantum hardware based on superconducting circuits which is among the most advanced and scalable platforms. The envisaged long-term solution of fault-tolerant quantum computers that correct their own errors remains out of reach mainly due to the associated qubit number overhead. As a result, the field has developed strategies that combine quantum and classical resources, exploit hardware-native operations, and employ error mitigation techniques to extract meaningful results from noisy data. This doctoral thesis contributes to this broader effort by exploring methods for advancing quantum simulation across the full computational stack, including hardware-level innovations, refined techniques for noise modeling and error mitigation, and algorithmic improvements enabled by efficient measurement processing.
Modern society increasingly depends on ever more capable computing devices. Yet some important problems remain beyond the reach of classical computers due to their inherent computational complexity. At the same time, the energy required to power this growing computational demand is an important concern. Over the past decade, quantum computers have received increased attention, not as a replacement for classical systems, but as a complementary technology with the potential to tackle complex problems more efficiently and significantly boost computational performance. This is made possible by leveraging the principles of quantum mechanics. To achieve this, quantum computers rely on a fundamental basic block called a qubit. There are several different technologies for realizing qubits, and no single approach has yet established itself as the definitive choice. In this work, we focus on silicon spin qubits in quantum dots, which use the intrinsic spin states of an electron as the computational basis states. This platform offers notable advantages including long coherence times, compatibility with the existing semiconductor industry and scalability. However, this technology is still very immature and substantial progress is still required. Just like for the design of classical electronics, simulations can play a crucial role by reducing development cycles, lowering costs, and improving overall efficiency. Throughout this thesis, we will propose a comprehensive simulation framework that allows to study and design spin qubit systems focusing on key aspects of the system architecture. First, a spin qubit compact model is presented that enables the co-simulation of spin qubits together with the classical CMOS control and readout circuitry that will significantly impact the performance of the qubits. We show that the model is capable of including all elements to build an elementary quantum gate set and illustrate its use by co-simulating a spin qubit coupled to a readout resonator with a complete cryo-CMOS readout chip design in a commercial electronic design automation tool. Next, we discuss a hybrid-Hamiltonian simulation approach that is aimed at simulating spin qubit microarchitectures that are suited for running quantum error correction protocols, which protect the encoded information against the errors that unavoidably happen due to the inherently noisy physical qubit devices. This simulation approach incorporates several novel strategies to tackle the challenge of the required computing power that grows exponentially with the size of the simulated system. Furthermore, additional functionality is incorporated to evaluate and track the performance of the implemented error correction protocols. Finally, we propose a characterization protocol for constructing a correlated noise model of the charge noise that is an important factor affecting spin qubit performance. The sources of charge noise are modelled as two-level fluctuators and the entire qubit array is characterized by dividing it into patches and running batches of benchmarking quantum gate sequences. The resulting noise model enables to scale the simulations to larger system sizes and to study the system performance and hardware demands of error correction protocols. Together, these models form a multi-layer simulation framework that enables detailed analysis and optimization of future spin qubit microarchitectures, with particular emphasis on error correction protocols. As such, the framework contributes to the realization of practical large-scale quantum computing.
Abstract As research on building scalable quantum computers advances, it is important to be able to certify their correctness. Due to the exponential hardness of classically simulating quantum computation, straight-forward verification through classical simulation fails. However, we can classically simulate small scale quantum computations and hence we are able to test that devices behave as expected in this domain. This constitutes the first step towards obtaining confidence in the anticipated quantum-advantage when we extend to scales that can no longer be simulated. Realistic devices have restrictions due to their architecture and limitations due to physical imperfections and noise. Here we extend the usual ideal simulations by considering those effects. We provide a general methodology for constructing realistic simulations emulating the physical system which will both provide a benchmark for realistic devices, and guide experimental research in the quest for quantum-advantage. We exemplify our methodology by simulating a networked architecture and corresponding noise-model; in particular that of the device developed in the Networked Quantum Information Technologies Hub (NQIT) (Networked Quantum Information Technologies Hub 2018 https://nqit.ox.ac.uk/ ; 2016 private communication. The error model was based on communication with Chris Balance and Tom Harty, mediated through Niel de Beaudrap, early on the NQIT project. Continued collaboration and communication with experimentalists could lead in refinement of the error model, which could be subject for future work.). For our simulations we use, with suitable modification, the classical simulator of Bravyi and Gosset 2016 ( Phys. Rev. Lett. 116 250501). The specific problems considered belong to the class of instantaneous quantum polynomial-time ( IQP ) problems (Shepherd and Bremner 2009 Proc. R. Soc. A 465 141339), a class believed to be hard for classical computing devices, and to be a promising candidate for the first demonstration of quantum-advantage. We first consider a subclass of IQP , defined in Bermejo-Vega et al 2018 ( Phys. Rev. X 8 021010), involving two-dimensional dynamical quantum simulators, before moving to more general instances of IQP , but which are still restricted to the architecture of NQIT.
Molecules are serious candidates for building hardware for quantum computers. They can encode quantum information onto electron or nuclear spins and some of them show important features as the scalability of the number of qubits and a universal set of quantum gates. In this paper we present our advances in the development of a classical simulation infrastructure for molecular Quantum Computing: starting from the definition of simplified models taking into account the main physical features of each analyzed molecule, quantum gates are defined over these models, thus permitting to take into account the real behavior of each technology during the simulation. An interface with a hardware-agnostic description language has been also developed. The knowledge of the behavior of real systems permits to optimize the design of quantum circuits at both physical and compilation levels. Elementary quantum algorithms have been simulated on three different molecular technologies by changing the physical parameters of polarization and manipulation and quantum circuit design strategies. Results confirm the dependency of the fidelity of the results on both levels, thus proving that the choice of optimal operating points and circuit optimization techniques as virtual-Z gates are fundamental for ensuring the execution of quantum circuits with negligible errors.