VQA-Classification-Algorithm-Oriented Performance Benchmarks for Quantum Computing

• Published in: 2023 8th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA) pp. 407 - 412
• Main Authors: Zhang, Meng, Wang, Jing, Yao, Fei, Lai, Junsen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 26.04.2023
• Subjects: Benchmark testing; Computers; data classification; Force; Parallel processing; performance benchmarks; Quantum algorithm; Quantum computing; Qubit; variational quantum algorithms
• ISSN: 2832-3734
• DOI: 10.1109/ICCCBDA56900.2023.10154749

Quantum computing, taking advantage of its parallel computing, is expected to provide exponential acceleration on some difficult problems. It is one of the important directions for the leapfrog development of computing power in the future. However, due to the limitation of the number of quantum bits, coherence time, fidelity and other factors, the computing power of quantum computers has not been fully utilized at the current stage. In recent years, researchers have proposed a variety of quantum computing performance benchmarks to evaluate and research the performance of quantum computers from the bit, circuit, system, application and other levels. However, the existing performance benchmarks cannot directly evaluate the ability of quantum computers to solve specific problems. Considering the above problems, this paper proposes a VQA-classification-algorithm-oriented performance benchmarks for quantum computing, which takes the classification effect under different data scales as one of the evaluation indicators. This benchmark can comprehensively evaluate the ability of quantum computers to solve data classification problems from the three dimensions of scale, speed, and quality. The benchmark was used to evaluate the superconducting quantum computer and simulator of the IBM and Huawei quantum platform, and its effect was verified experimentally.