Digital quantum computers promise exponential speedups in performing quantum time-evolution, providing an opportunity to simulate quantum dynamics of complex systems in physics and chemistry. However, the task of extracting desired quantum properties at intermediate time steps remains a computational bottleneck due to wavefunction collapse and no-fast-forwarding theorem. Despite significant progress towards building a Fault-Tolerant Quantum Computer (FTQC), there is still a need for resource-efficient quantum simulators. Here, we propose a hybrid simulator that enables classical computers to leverage FTQC devices and quantum time propagators to overcome this bottleneck, so as to efficiently simulate the quantum dynamics of large systems initialized in an unknown superposition of a few system eigenstates. It features no optimization subroutines and avoids barren plateau issues, while consuming fewer quantum resources compared to standard methods when many time steps are required.
Digital quantum computers promise exponential speedups in performing quantum time-evolution, providing an opportunity to simulate quantum dynamics of complex systems in physics and chemistry. However, the task of extracting desired quantum properties at intermediate time steps remains a computational bottleneck due to wavefunction collapse and no-fast-forwarding theorem. Despite significant progress towards building a Fault-Tolerant Quantum Computer (FTQC), there is still a need for resource-efficient quantum simulators. Here, we propose a hybrid simulator that enables classical computers to leverage FTQC devices and quantum time propagators to overcome this bottleneck, so as to efficiently simulate the quantum dynamics of large systems initialized in an unknown superposition of a few system eigenstates. It features no optimization subroutines and avoids barren plateau issues, while consuming fewer quantum resources compared to standard methods when many time steps are required.
The accurate and efficient simulation of many body systems has been a long-standing challenge for quantum chemists and physicists. At the heart of this challenge lies the correlation between quantum particles, which is crucial in understanding various physical phenomena, but is numerically hard to calculate. Over the decades, there has been remarkable progress in simplifying this problem aided by mathematical and chemical approximations and the availability of powerful computers. However, an exact solution to the problem still remains intractable as it requires exponential re- sources with respect to the system size. Recent advances in the field of quantum computing have shown promise in Feynman’s idea of simulating quantum mechanics on quantum computers. It has been proposed that an ideal fault-tolerant quantum computer can reduce the scaling of these sim- ulations from exponential to polynomial. However, initial quantum algorithms have shown that there is a large gap between the capabilities of current hardware and the resources required for simulating quantum systems of interest. The work of this thesis focuses on reducing this gap by introducing two novel algorithms for quantum xiv chemistry simulations, which are suitable for near-term hardware. First, we intro- duce a hybrid algorithm to decrease the total number of operations required for the quantum simulations. We use a classical computer to generate an effective Hamilto- nian containing O(n2) terms compared to the O(n4) terms of the full Hamiltonian, where n is the number of orbitals. This sparse Hamiltonian can then be used with a high-level method to recover the ground state energies. We demonstrate that this sparsification of the Hamiltonian reduces the number of quantum operations required for the simulation by an order of magnitude, thus making it accessible for the near- term Hardware. Our second algorithm aims to make the calculation of dynamic correlation functions (Green’s functions) more feasible on these quantum machines. This algorithm is designed with the motive of avoiding the use of time-evolution and reduced use of two-qubit gates. We demonstrate that we can reproduce the Green’s function within reasonable error limits using this approach. We have also discussed the use of re-sampling techniques for proper error-propagation of the stochastic data obtained from the quantum machines. The general structure of this thesis is as follows. We begin by motivating this work in Chapter 1. Chapters 2 and 3 provide a general introduction of quantum computing and Green’s function based methods, respectively. In chapter 4, the development of effective Hamiltonian is discussed in detail. Chapter 5 demonstrates the use of these effective Hamiltonians on the quantum machine for time evolution. In chapter 6, we have discussed the algorithm to calculate Green’s functions on a quantum machine. This is followed by conclusions and future directions in Chapters 7 and 8, respectively.
Estimating the eigenvalue or energy gap of a Hamiltonian H is vital for studying quantum many-body systems. Particularly, many of the problems in quantum chemistry, condensed-matter physics, and nuclear physics investigate the energy gap between two eigenstates. Hence, how to efficiently solve the energy gap becomes an important motive for researching new quantum algorithms. In this work, we propose a hybrid nonvariational quantum algorithm that uses the Monte Carlo method and real-time Hamiltonian simulation to evaluate the energy gap of a general quantum many-body system. Compared to conventional approaches, our algorithm does not require controlled real-time evolution, thus making its implementation much more experimental friendly for near-term quantum devices and early fault-tolerant quantum computers. To verify the efficiency of our algorithm, we conduct numerical simulations for the Heisenberg model and molecule systems on a classical emulator.
This paper introduces an algorithmic quantum simulation framework designed to efficiently model and understand complex systems, particularly those exhibiting emergent properties. The core aim is to leverage the strengths of both quantum and classical computation to accelerate simulation and analysis of these systems. We propose a hybrid approach that utilizes quantum algorithms for key computational bottlenecks, alongside classical optimization and exploration techniques to achieve a significant speedup over traditional simulation methods. This work centers on a specific example—modeling the dynamics of biological networks – and demonstrates the potential for this hybrid strategy to unlock new insights into complex phenomena. The paper concludes with a discussion of the challenges and future directions of this research.
Hybrid computing models, which integrate both classical and quantum systems, represent a promising frontier in enhancing computational power and efficiency. Classical computing systems, while highly efficient for a broad range of tasks, encounter significant limitations when addressing complex, large-scale problems such as optimization, cryptography, and molecular simulations. Quantum computing, on the other hand, offers exponential speedup for specific algorithms, particularly in areas like factorization, search, and simulation of quantum systems. However, quantum systems are not yet mature enough to replace classical systems entirely, due to issues like error rates, qubit coherence, and the need for cryogenic environments. As a result, hybrid models are emerging as the most viable solution, combining the strengths of both paradigms to tackle computationally intensive tasks more efficiently. This paper provides a comprehensive analysis of the development and application of hybrid computing models. It explores how these models leverage the parallel processing power of quantum systems while utilizing the robustness and data handling capabilities of classical systems. We discuss various architectures proposed for hybrid computing, focusing on their scalability, efficiency, and real-world applications across industries such as healthcare, finance, cryptography, and artificial intelligence. Additionally, we examine the challenges involved in integrating classical and quantum systems, including synchronization, resource management, and algorithm development. The paper concludes with future directions for hybrid computing, emphasizing the potential for these models to revolutionize fields that rely on high-performance computing by offering unprecedented computational capabilities. In summary, hybrid computing models stand at the intersection of classical and quantum technologies, providing a synergistic approach to solving some of the most challenging computational problems faced today.
Quantum computing offers promising potential for faster computation compared to classical computing, with the advantage of improved scaling as the problem size increases.Many fast quantum algorithms have been developed to solve specific problems using quantum computation, and they are expected to be applied to a wide range of fields, including quantum chemistry, condensed matter physics, and machine learning.However, the high cost of realizing physical systems to perform quantum computations makes resource-efficient implementation methods indispensable.In the first part of this thesis, we study the resource-efficient design of a fault-tolerant protocol based on quantum low-density parity-check (LDPC) codes.Scalable quantum computation to achieve the speed-ups over classical computation requires fault tolerance due to the vulnerability of quantum systems to noise.Fault-tolerant quantum computation (FTQC) expenses significant space and time overheads to achieve fault-tolerant quantum circuits from the original circuits intended to simulate.An FTQC protocol based on quantum LDPC codes has the attractive advantage of constant-space overhead, but has the disadvantage of requiring a long polynomial-time overhead.In this thesis, we present a polylogarithmic time overhead FTQC protocol based on quantum LDPC code, which is exponentially time efficient compared to conventional protocols while maintaining the advantage of the constant space overhead.This overhead is enjoyed even if the non-zero classical computation performing the decoding algorithms is considered.The second part of this thesis investigates the resource-efficient design of variational quantum algorithms (VQAs).The VQAs are a heuristic algorithm with promising applications in various fields.The VQAs can be performed by minimizing task-specific cost functions, typically defined as the expectation values of observables.However, it is known that the execution of variational quantum algorithms requires numerous quantum measurements, and their reduction is a significant challenge in algorithm implementation.In the first study of this part, we propose a new algorithm for calculating nonadiabatic couplings (NACs), which are important quantities in simulating molecular
Parallel Quantum Simulation is a rapidly evolving field within quantum computing, aiming to overcome the limitations of single-qubit or small-system simulations for complex quantum systems. This paper explores a novel hybrid quantum-classical algorithm designed to significantly enhance the computational efficiency of quantum simulations. The core mechanism centers around intelligently combining quantum algorithms for precise calculations with classical algorithms for data processing and control, optimizing the overall simulation process. We present a detailed algorithmic framework, outlining the key steps involved, and discuss the potential benefits, including increased accuracy, reduced computational cost, and improved scalability. This research addresses a critical bottleneck in quantum computing, offering a pathway towards more realistic simulations and accelerating the development of practical quantum applications.