Research On Large Scale Quantum Computing Simulation Technology

Quantum computing is a new type of computing simulation,a subject that intersects quantum theory and computing science,and is the latest development trend of current computing technology.Due to its use of superposition,entanglement and reversibility,it has unparalleled advantages in parallel computing.However,the current quantum computer capable of exerting huge computing power has not been successfully developed.Due to the limitation of the technological level of physical realization,it takes time and technical accumulation to solve practical problems of practical significance.Therefore,the quantum computing simulation technology studied in this article has important practical value and is an important means and effective method for carrying out quantum computing theoretical research and application exploration.The research contents of this paper are as follows:1.Implementation of HHL algorithm deployment on the original quantum platform.Through in-depth research on the classic HHL algorithm,a small-scale quantum circuit model is proposed to solve a 2×2 linear equation system.Two methods of quantum circuit diagram and advanced quantum programming are used for simulation verification on the original quantum platform.When b takes different parameter values,the fidelity reaches more than 90%.2.Propose and implement the "Classical + Quantum" quantum simulation framework.At present,quantum computer hardware and simulation platforms are not enough to provide enough qubits to fully simulate existing quantum algorithms.Therefore,this article relies on supercomputers to propose and implement a "Classical + Quantum" quantum simulation framework to achieve super-large-scale quantum algorithms simulation.This article uses the existing high-level programming language(C language)to abstract the definition of quantum logic operations,and simulates the QFT algorithm on the "Sunway Taihu Light" supercomputer and the Shor algorithm on the "Songshan" supercomputer.3.On "Shenwei Taihu Light",the QFT algorithm was used to verify the quantum simulation framework of "Classical + Quantum".Relying on the master-slave architecture of the Shenwei platform,a high-level programming language(C language)is used to abstract the quantum circuit definition of the QFT algorithm.According to the characteristics of its heterogeneous many-cores,the two-level parallel strategy of "multi-process MPI parallel + in-process master-slave parallel" and the method of communication and calculation hiding are optimized to shorten the simulation time of quantum computing and improve simulation efficiency.And use the script to achieve the purpose of semi-automatic execution of the algorithm.Finally,the 46-qubits QFT algorithm was successfully deployed on "Sunway Taihu Light",and the speedup ratio and MPI parallel efficiency were good.4.On "Songshan",the "Classical + Quantum" quantum simulation framework was verified through the simulation of Shor's algorithm.By using the "Songshan" supercomputer system,a high-level programming language(C language)is used to abstract the quantum circuit definition of the Shor algorithm.It is optimized according to the CPU+DCU heterogeneous cluster architecture and the characteristics of the Shor algorithm.The algorithm is cleverly divided into two parts for implementation.The MPI+HIP programming model is used to simulate the inverse QFT module in the CPU+DCU heterogeneous cluster,which improves the simulation efficiency of the entire algorithm.Finally,the 42-qubits Shor algorithm was successfully deployed on "Songshan".