IBM Quantum Job Scheduling

Mon, 01 Jun 2020 00:00:00 +0000

Designed and implemented the fair-share queue that decides the order in which jobs run on IBM Quantum hardware. It balances access across users and organizations according to their allocations, and it schedules every job run on every IBM Quantum QPU — still in production today.

Later, as Data & Intelligence Chapter Lead, I led work on predicting QPU processing time for user jobs, which has been deployed on production since 2023. Based on this experience, we published a paper at IEEE QCE 2025 in which we used a dataset of ~150,000 jobs following the IBM Quantum schema, gradient-boosting models (LightGBM) forecast how long each job will take on the hardware, giving scheduling and resource management a better signal than static estimates.

Production Systems | Juan Cruz-Benito

IBM Quantum Job Scheduling