QuaST
Quantum-enabling Services and Tools for Industrial Applications
Project description
Quantum computing can lead to disruptive changes in many areas of industry, for example by solving complex optimization problems in logistics or semiconductor manufacturing more efficiently. However, in order for industry and society to fully benefit from future quantum computers, a low-threshold access is needed, which in particular does not require the user to have a deep knowledge of the underlying physics. QuaST will make this possible by developing software and tools in a unique holistic approach.
Challenge and innovation
Designing quantum algorithms is unintuitive to both classical programmers and chip designers even when they are embedded in the API of widely used high-level languages such as Python. In particular, most software development steps currently still have to be done in a hardware-sensitive manner: What hardware is best suited for a specific application problem (e.g., what chip layout)? How can the quantum algorithm best be implemented on the native architecture of the chosen quantum computing hardware? How is it guaranteed that the quantum computer will produce robust and reliable results?
The QuaST project offers a unique, comprehensive approach that will require only minimal knowledge of the quantum computing hardware and hardware-related software components from the end user in industry. The goal is to provide the end-user with high-level libraries that, based on the application problem, automatically decompose the solution into classical, HPC and quantum computing parts, and then map these quantum computing parts optimized to the hardware level, including through co-design.
Solution approach
In order to solve complex optimization problems in a holistic approach with as little prior knowledge of the end user as possible, QuaST develops high-level libraries and services. The development of these high-level libraries comprises a software stack starting from the optimization problems of the application partners down to the co-design level. For this purpose, methods are developed to implement the posed optimization problems optimally on IQM hardware in a co-design strategy, and to connect the IQM hardware to the ParityOS operating system in the process. In addition, high-level solution paths in the form of a library of algorithmic building blocks and high-level mapping techniques will be developed to automatically translate the application algorithms to different quantum computing hardware using the ParityOS operating system. In addition, further tools will be provided to suitably decompose the application problems into classical, HPC-accelerated, and quantum computing-based parts, or into smaller subproblems that can make optimal use of the current NISQ hardware. In this context, methods for (de)encoding the classical input information are developed, as well as tools for optimization, validation, and performance analysis. Another work package optimizes the runtime environment between classical computers and quantum computers to execute quantum algorithms in a runtime-efficient manner. To ensure that quantum computing hybrid algorithms can be used reliably, methods to evaluate and verify the application software are also being developed.
Consortium
Fraunhofer IKS (project lead), TU München, Leibniz Rechenzentrum, IQM, ParityQC, Infineon, DATEV eG, Fraunhofer AISEC, Fraunhofer IIS, Fraunhofer IISB
Duration
January 2022 – December 2024
Budget
Total costs: € 7.7 million
Funding volume: € 5.5 million