Quantum Program Loop Scheduling to Cut Hybrid Execution Delays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hybrid quantum-classical algorithms experience prolonged execution times due to frequent transitions between quantum and classical computers, leading to significant waiting times and inefficient resource utilization, particularly in variational quantum algorithms (VQAs), which can take over 100 days to complete with current job scheduling methods.
Innovation Solution
A method to determine a loop process section in a quantum program, estimate execution and utilization times of a quantum computer within that section, and schedule jobs based on these estimates to prioritize and optimize the use of quantum resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If job scheduling is performed using conventional methods with frequent transitions between quantum and classical computers, then the algorithm can be executed, but the execution time becomes excessively long (over 100 days)
Solution Approach 1:
The patent applies preliminary action by estimating the quantum computer utilization time in advance before actual job execution. The job scheduler uses this pre-estimated time to proactively schedule subsequent jobs and optimize resource allocation, rather than reacting to actual execution times after transitions occur. This allows the system to prepare and minimize waiting periods beforehand, significantly reducing the overall execution time of hybrid quantum-classical algorithms.
2Reliability
If the quantum computer is frequently utilized in loop processes, then the algorithm accuracy can be improved, but the resource utilization becomes inefficient due to repeated waiting
Solution Approach 1:
The patent implements feedback by measuring actual quantum computer utilization times during loop process executions and using this information to refine future time estimates. The job scheduler continuously adjusts scheduling decisions based on feedback from previous executions, optimizing the balance between achieving sufficient algorithm accuracy through multiple loop iterations and minimizing resource waste from repeated waiting periods. This creates a closed-loop system that improves both reliability and productivity over time.
Data Source
AI summary
A non-transitory computer-readable recording medium stores therein an information processing program that causes a computer to execute a process including acquiring a call stack of a first program for utilizing a quantum computer, determining a section of a loop process in the first program, based on the call stack, measuring a first execution time of the section determined and a first utilization time of the quantum computer in the section determined, and estimating, based on the first execution time and the first utilization time, a second execution time of the section determined and a second utilization time of the quantum computer in the section determined.