Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bridging HPC and Quantum Systems using Scientific Workflows

Samuel T. Bieberich, Ketan C. Maheshwari, Sean R. Wilkinson, Prasanna Date, In-Saeng Suh, Rafael Ferreira da Silva

TL;DR

This work addresses the challenge of enabling practical hybrid computing by bridging traditional HPC resources with quantum processors through scientific workflows. It demonstrates an end-to-end HPC–quantum workflow, orchestrating three quantum algorithms—Grover’s search, Shor’s factoring, and a 4-node Traveling Salesman Problem—over IBMQ using Parsl. The results show that simulators offer high fidelity while real quantum hardware remains noisy for larger circuits, highlighting both the promise and current limitations of quantum accelerators in hybrid workflows. The contributions include a general, workflow-agnostic bridging approach and a concrete demonstration across multiple algorithms, underscoring the potential for HPC centers to integrate quantum resources as auxiliary processors and setting the stage for broader platform support in future work.

Abstract

Quantum Computers offer an intriguing challenge in modern Computer Science. With the inevitable physical limitations to Moore's Law, quantum hardware provides avenues to solve grander problems faster by utilizing Quantum Mechanical properties at subatomic scales. These futuristic devices will likely never replace traditional HPC, but rather work alongside them to perform complex tasks, utilizing the best of decades of HPC and quantum computing research. We leverage the capabilities of scientific workflows to make traditional HPC and Quantum Computers work together. To demonstrate this capability, we implemented three algorithms: Grover's Search Algorithm, Shor's Factoring Algorithm, and a 4-node Traveling Salesman Algorithm. The algorithms' implementation and generated inputs are sent from ORNL HPC to IBMQ, the algorithms run on IBMQ, and the results return. The entire process is automated as a workflow by encoding it into the Parsl parallel scripting and workflow platform.

Bridging HPC and Quantum Systems using Scientific Workflows

TL;DR

This work addresses the challenge of enabling practical hybrid computing by bridging traditional HPC resources with quantum processors through scientific workflows. It demonstrates an end-to-end HPC–quantum workflow, orchestrating three quantum algorithms—Grover’s search, Shor’s factoring, and a 4-node Traveling Salesman Problem—over IBMQ using Parsl. The results show that simulators offer high fidelity while real quantum hardware remains noisy for larger circuits, highlighting both the promise and current limitations of quantum accelerators in hybrid workflows. The contributions include a general, workflow-agnostic bridging approach and a concrete demonstration across multiple algorithms, underscoring the potential for HPC centers to integrate quantum resources as auxiliary processors and setting the stage for broader platform support in future work.

Abstract

Quantum Computers offer an intriguing challenge in modern Computer Science. With the inevitable physical limitations to Moore's Law, quantum hardware provides avenues to solve grander problems faster by utilizing Quantum Mechanical properties at subatomic scales. These futuristic devices will likely never replace traditional HPC, but rather work alongside them to perform complex tasks, utilizing the best of decades of HPC and quantum computing research. We leverage the capabilities of scientific workflows to make traditional HPC and Quantum Computers work together. To demonstrate this capability, we implemented three algorithms: Grover's Search Algorithm, Shor's Factoring Algorithm, and a 4-node Traveling Salesman Algorithm. The algorithms' implementation and generated inputs are sent from ORNL HPC to IBMQ, the algorithms run on IBMQ, and the results return. The entire process is automated as a workflow by encoding it into the Parsl parallel scripting and workflow platform.
Paper Structure (14 sections, 2 equations, 8 figures)

This paper contains 14 sections, 2 equations, 8 figures.

Table of Contents

  1. Introduction
  2. Traveling Salesman Problem
  3. Algorithm
  4. Implementation
  5. HPC-Quantum Workflow
  6. Results
  7. Grover's Search Algorithm
  8. Implementation
  9. Results
  10. Shor's Factoring Algorithm
  11. Implementation
  12. Results
  13. Related Work
  14. Conclusions and Future Work

Figures (8)

  • Figure 1: Overall workflow schema between tradional HPC and Quantum Systems.
  • Figure 2: Example NetworkX map for a four-node TSP.
  • Figure 3: Circuit for 4 node TSP with Eigenstate 11000110, or path 1-2-3-4-1
  • Figure 4: 1-2-3-4 path TSP circuit results from ibmq_kolkata. Correct answer highlighted.
  • Figure 5: IBMQ Matplotlib circuit for Grover's Algorithm, with one implementation of the Grover Operator. Value being searched for is 15.
  • ...and 3 more figures