Alibaba Cloud Quantum Development Platform: Applications to Quantum Algorithm Design

Cupjin Huang,Mario Szegedy,Fang Zhang,Xun Gao,Jianxin Chen,Yaoyun Shi
DOI: https://doi.org/10.48550/arXiv.1909.02559
2019-09-06
Abstract:We report our work on the Alibaba Cloud Quantum Development Platform (AC-QDP). The capability of AC-QDP's computational engine was already reported in \cite{CZH+18, ZHN+19}.In this follow-up article, we demonstrate with figures how AC-QDP helps in testing large-scale quantum algorithms (currently within the QAOA framework). We give new benchmark results on regular graphs. AC-QDP's QAOA framework can simulate thousands of qubits for up to $4$ layers. Then we discuss two interesting use cases we have implemented on the platform: 1. Optimal QAOA sequences for small-cycle free graphs; 2. Graph structure discovery.
Quantum Physics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: how to use Alibaba Cloud Quantum Development Platform (AC - QDP) to assist in the testing and optimization of large - scale quantum algorithms, especially within the framework of Quantum Approximate Optimization Algorithm (QAOA). Specifically, the paper focuses on the following aspects: 1. **Simulation of large - scale quantum algorithms**: - The paper shows how AC - QDP can simulate QAOA circuits containing thousands of qubits and up to 4 - layer depth through classical computing resources. - Through new benchmark test results, the performance of AC - QDP on regular graphs is proven. 2. **Application of QAOA on specific graph structures**: - **Regular graphs without small cycles**: The optimal sequence of QAOA on regular graphs without small cycles is studied, and the influence of these graph structures on the performance of QAOA is discussed. - **Graph structure discovery**: The application of QAOA in identifying graph isomorphism problems is explored, and a method based on the QAOA energy function is proposed to distinguish non - isomorphic graphs. 3. **Application of optimization techniques**: - Multiple global and local optimization techniques (such as differential evolution, FOURIER heuristic, etc.) are used to optimize the angle parameters in QAOA to improve the performance of the algorithm. - Especially when dealing with graphs with larger degrees and layers, the effectiveness of the optimization strategy is demonstrated. ### Specific problem description - **Limitations and improvements of QAOA**: QAOA cannot fully simulate the quantum adiabatic algorithm, but by introducing freely optimizable parameters, it becomes a hybrid quantum - classical algorithm. The paper explores how to find the optimal values of these parameters through classical optimization methods, thereby improving the performance of QAOA. - **Challenges of classical simulation**: Due to the extremely high memory requirements for classical storage of quantum states, the traditional state - vector simulation method becomes infeasible when dealing with more than about 50 qubits. The paper proposes a tensor - network - based simulation method, which can effectively meet this challenge and handle larger - scale quantum systems. - **Practical application scenarios**: The paper also shows the potential of AC - QDP in practical application scenarios, such as its application in graph isomorphism problems and its performance in combinatorial optimization problems (such as MAX - CUT). ### Summary The main objective of the paper is to show the capabilities and advantages of AC - QDP in the testing and optimization of large - scale quantum algorithms, especially through the QAOA framework, to solve the curse - of - dimensionality problem faced by traditional simulation methods, and to provide strong support for the future development of quantum computing.