Exploiting Multilevel Parallelism in Quantum Simulations

B. Paredes, Frank Verstraete, J. I. Cirac

We present an algorithm that exploits quantum parallelism to simulate randomness in a quantum system. In our scheme, all possible realizations of the random parameters are encoded quantum mechanically in a superposition state of an auxiliary system. We show how our algorithm allows for the efficient simulation of dynamics of quantum random spin chains with known numerical methods. We propose an experimental realization based on atoms in optical lattices in which disorder could be simulated in parallel and in a controlled way through the interaction with another atomic species.

Mark Richter, Binh Trieu, G. W. Arnold +1 more

Karen Arzumanyan, Roman Sahakyan, Romik Sargsyan +3 more

Abstract Simulating multilevel quantum systems using the density matrix formalism involves solving large systems of coupled differential equations, often requiring computationally intensive methods such as Runge–Kutta, Monte Carlo, or physics-informed neural networks. These solvers demand high-performance computing resources because of the complexity and size of the underlying matrices. This study evaluates key aspects of simulation performance, such as numerical accuracy, scalability, parallel efficiency, or resource utilization, to develop efficient computational strategies. To address infrastructure overhead and improve accessibility, a Function-as-a-Service-based platform is proposed to enable elastic, on-demand execution of quantum simulation workflows. A monitoring and orchestration strategy is integrated into the platform to optimize resource use across varying problem dimensionalities. By abstracting the complexity of traditional high-performance computing environments, the proposed Function-as-a-Service-based approach facilitates efficient and scalable quantum simulations while reducing user-side infrastructure management. The results provide a foundation for designing accessible, high-fidelity simulations for complex quantum systems on modern cloud-based or hybrid high-performance computing platforms.

Koen De Raedt, Kristel F. L. Michielsen, Hans A. De Raedt +6 more

Frank Tabakin A, Bruno Julia Diaz

Zhikun Han, Chufan Lyu, Yuxuan Zhou +17 more

Energy spectroscopy is a powerful tool with diverse applications across various disciplines. Variational quantum-classical algorithms based on programmable digital quantum simulators have emerged as promising approaches for conducting spectroscopy on various models using a single device, despite facing significant quantum and classical resource overheads. Here, we experimentally demonstrate multilevel variational spectroscopy for fundamental many-body Hamiltonians using a superconducting digital quantum simulator. By exploiting symmetries and the subspace search method, we achieve full spectroscopy for a four-qubit Heisenberg spin chain, yielding an average energy deviation as small as 0.13 from the theoretical values, assuming unity coupling strength. Our method, when extended to eight-qubit Heisenberg and transverse-field Ising Hamiltonians, successfully determines the three lowest-energy levels. In achieving the above, we introduce a circuit-agnostic compilation method that enhances the robustness of our simulator against signal crosstalk. Our study highlights the combination of the subspace search method and symmetry-assisted resource efficiency in variational quantum algorithms and lays the foundation for practical spectroscopy on near-term quantum simulators, with potential applications in quantum chemistry and condensed matter physics. Published by the American Physical Society 2024