Accelerating Randomized Benchmarking using Conditional Statements on Quantum Cloud Computing

양자 클라우드 컴퓨팅에서 조건문을 사용한 Randomized Benchmarking 가속화

Randomized benchmarking (RB) techniques for measuring the error rate of quantum computers require a large number of sequences and repeated experiments for accurate error characterization. The cloud quantum computing service, IBM Quantum Platform, offered by IBM, utilizes a first-in, first-out (FIFO) job scheduler for the purpose of ensuring fair resource distribution. However, this scheduling policy is subject to a structural limitation, in that each sequence is queued and executed individually. In this paper, a method is proposed for integrating RB sequences composed of multiple sequences into a single task, with the objective of minimizing the queue waiting time. Experiments conducted on the IBM Quantum Platform demonstrated that the proposed method could effectively reduce the overall experimental time by 70.9% and decrease the pending overhead by 64.9% compared to conventional methods. Additionally, an efficient error characterization solution was proposed under the constraints of cloud quantum computing environments.