Software for Hybrid Quantum Computing

Integrating quantum into supercomputing is not just about hardware installations. The operating software of both systems must also converge.

decorative image

decorative image

If you want to use quantum computing today, you need supercomputing capabilities. Classical and quantum processors process each other's jobs, and data flows between the two systems. To achieve this, quantum systems must be deeply integrated into high-performance computing (HPC), which is not only a question of hardware but also affects software development. Specialists from the Leibniz Supercomputing Centre (LRZ) have recently published a paper entitled "Accelerating HPC with Quantum Computing: It's a Software Challenge Too" in the journal Computing in Science and Engineering of the Institute for Electrical and Electronic Engineering (IEEE): "This article describes our vision for an integrated ecosystem that brings together existing HPC and evolving quantum software stacks into a single system to enable a common and continuous user experience," it states, and: "Quantum systems present different requirements (...). To meet them, new scheduling approaches on the HPC side and new programming approaches on the QC side will be required".

The Quantum Computing and Technologies (QCT) department at the LRZ is working in various partnerships to extend existing software stacks in order to link the two technologies more closely. The problem: different programme characteristics. Supercomputer software aims to optimise execution time, and programmers use Fortran, C or C++ to achieve this. These languages integrate as many tasks as possible into the compilation, optimisation and job preparation, and transform the code into machine language so that it runs efficiently. Typical quantum computing programming languages, Qiskit or Cirq,  are based on Python and define circuits to solve problems. A separate programme, the interpreter, translates the commands into machine language. The article describes how HPC can be used to parallelize the generation and optimisation of quantum circuits, which tasks compilers and optimisers should solve, and how the flow of data between systems can be accelerated using the Rust programming language.
 
Martin Schulz, Martin Ruefenacht, Laura Brandon Schulz, Dieter Kranzlm├╝ller Accelerating HPC with quantum computing: It's Also a Software Challenge. From: Computing in Science and Engineering, IEEE Computer Society, 2023 

 

Get Connected: LRZ in Social Media

Leibniz Supercomputing Centre
of the Bavarian Academy of Sciences and Humanities

Boltzmannstra├če 1
85748 Garching near Munich

Phone: +49(0)89 - 35831 8000
Email: presse@lrz.de
Website: www.lrz.de