Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantum Processing Unit (QPU) processing time Prediction with Machine Learning

Lucy Xing, Sanjay Vishwakarma, David Kremer, Francisco Martin-Fernandez, Ismael Faro, Juan Cruz-Benito

TL;DR

The paper addresses predicting QPU processing time for quantum jobs to improve scheduling and resource management in quantum computing systems. It adopts a LightGBM gradient-boosting approach with careful data encoding on a large IBM Quantum dataset, and compares ML predictions to a heuristic baseline. The key contributions include demonstrating that ML predictions are more accurate than the heuristic for both sampler and estimator runtime primitives and providing a practical data-processing and modeling workflow. The findings suggest ML-enabled tooling can enhance quantum workload management and lay groundwork for integrating AI-driven features into quantum computing stacks.

Abstract

This paper explores the application of machine learning (ML) techniques in predicting the QPU processing time of quantum jobs. By leveraging ML algorithms, this study introduces predictive models that are designed to enhance operational efficiency in quantum computing systems. Using a dataset of about 150,000 jobs that follow the IBM Quantum schema, we employ ML methods based on Gradient-Boosting (LightGBM) to predict the QPU processing times, incorporating data preprocessing methods to improve model accuracy. The results demonstrate the effectiveness of ML in forecasting quantum jobs. This improvement can have implications on improving resource management and scheduling within quantum computing frameworks. This research not only highlights the potential of ML in refining quantum job predictions but also sets a foundation for integrating AI-driven tools in advanced quantum computing operations.

Quantum Processing Unit (QPU) processing time Prediction with Machine Learning

TL;DR

The paper addresses predicting QPU processing time for quantum jobs to improve scheduling and resource management in quantum computing systems. It adopts a LightGBM gradient-boosting approach with careful data encoding on a large IBM Quantum dataset, and compares ML predictions to a heuristic baseline. The key contributions include demonstrating that ML predictions are more accurate than the heuristic for both sampler and estimator runtime primitives and providing a practical data-processing and modeling workflow. The findings suggest ML-enabled tooling can enhance quantum workload management and lay groundwork for integrating AI-driven features into quantum computing stacks.

Abstract

This paper explores the application of machine learning (ML) techniques in predicting the QPU processing time of quantum jobs. By leveraging ML algorithms, this study introduces predictive models that are designed to enhance operational efficiency in quantum computing systems. Using a dataset of about 150,000 jobs that follow the IBM Quantum schema, we employ ML methods based on Gradient-Boosting (LightGBM) to predict the QPU processing times, incorporating data preprocessing methods to improve model accuracy. The results demonstrate the effectiveness of ML in forecasting quantum jobs. This improvement can have implications on improving resource management and scheduling within quantum computing frameworks. This research not only highlights the potential of ML in refining quantum job predictions but also sets a foundation for integrating AI-driven tools in advanced quantum computing operations.
Paper Structure (16 sections, 1 equation, 5 figures)

This paper contains 16 sections, 1 equation, 5 figures.

Table of Contents

  1. Introduction
  2. Materials and methods
  3. Materials
  4. Data Collection
  5. Model
  6. Hardware and software used
  7. Methods
  8. Encoding of the data
  9. Processing of the data
  10. Training of the model
  11. Predicting QPU time using an heuristic method
  12. Results
  13. Predicting QPU time
  14. QPU Time Prediction Prediction Accuracy
  15. Calculation of the multiplicative safety factor
  16. ...and 1 more sections

Figures (5)

  • Figure 1: Actual QPU time taken (blue line) vs. ML (orange) and heuristic (green) methods. Orange and green solid lines display moving averages over 50 points while individual predictions are shown as dots.
  • Figure 2: An equivalent set of colors and line formats are used here as in figure \ref{['fig:sampler_prediction']}.
  • Figure 3: QPU time prediction accuracy for sampler jobs using ML versus heuristic prediction methods.
  • Figure 4: QPU time prediction accuracy for estimator jobs using ML versus heuristic prediction methods.
  • Figure 5: Percentage of overestimated jobs after applying a particular multiplicative safety factor to its original prediction.