Osaka University and Partners Develop 'Quantum Multiprogramming (Auto Mode)' to Boost Quantum Computing Cloud Efficiency

## Highlights of Research and Development
- Developed a 'Quantum Multiprogramming (Auto Mode)' feature that automatically executes quantum programs from different users in parallel.
- The system selects optimal combinations from the job queue, effectively utilizing idle space on quantum chips.
- Drastically reduces waiting times and maximizes the utilization efficiency of the quantum computer cloud service operated by QIQB at Osaka University.
- Implemented in the open-source quantum computer base software 'OQTOPUS' and is being released to participating institutions of the 'Quantum Software Consortium.'

## Overview
Researchers at the Quantum Information and Quantum Biology Center (QIQB), Osaka University, have successfully developed and launched a 'Quantum Multiprogramming (Auto Mode)' feature. This function allows for the automatic parallel execution of quantum programs from various users within the cloud service environment.

This outcome enables the system to automatically select jobs from the cloud's job queue and assign them to available qubits for parallel execution. This represents a significant improvement over previous manual methods. It contributes to alleviating congestion and enhances the operating efficiency of domestic quantum computers by minimizing the underutilization of precious qubit resources.

## Realization of Auto Mode via Advanced Algorithms
- Optimal placement using graph theory: The system formulates quantum circuits and chip structures as 'graphs,' implementing a decision algorithm that fits multiple circuit graphs into the chip graph like a puzzle.
- Automatic resolution of physical constraints (Transpile function): Automatically converts quantum circuits considering hardware-specific constraints such as connection directions and connectivity between distant qubits.
- Job priority control considering fairness: Efficiently realizes parallelization by referring to jobs at the beginning of the queue and finding executable combinations while ensuring FIFO fairness.

## Impact on Society
Evaluation using datasets based on actual usage trends demonstrated that throughput for small-scale quantum circuits improved by approximately 3.76 times. As quantum chip bit counts increase, further improvements in efficiency are expected.