Quantum Computing Use of Quantum Computers in Industrial Simulation

Related Vendors

Fraunhofer-Institut für Integrierte Schaltungen IIS

FAULHABER Drive Systems

Wuxi InfiMotion Technology Co., Ltd.

Quantum computers in practical use in industry: a joint project aims to gain insights into this, as the potential is high. They could accelerate industrial process optimization and automation, as well as optimize energy distribution in smart grids.

The joint project "Q-Genesys—Quantum Generative Models for Industrial Simulation Systems" brings together experts from the Fraunhofer Institute for Integrated Circuits IIS, IQM Germany GmbH, Ostbayerische Technische Hochschule OTH Regensburg (Regensburg) and Siemens AG to harness the enormous technological potential of quantum computer systems for practical applications. The new findings will then be used for highly complex industrial applications, e.g. for the creation of complicated 3D designs, the design of new molecules or other chemical products.

The aim of the Q-Genesys project is to harness the supremacy of quantum computers for practical use in industry. This enables novel modeling and evaluation methods for a wide range of industrial applications. These range from the creation of new molecular structures for chemistry, pharmacy and medicine to industrial process optimization and automation. The project is scheduled to run for three years and will also focus on optimizing energy distribution for network management in smart grids and controlling building automation.

Making Highly Complex Benchmarking Faster And More Efficient

The Federal Ministry of Research, Technology and Space BMFTR is funding this project in the quantum systems program "Application-oriented quantum informatics" in order to combine quantum computing and generative learning and make them usable for more efficient, new product developments.

The Q-Genesys consortium partners are committed to closing the gap between conventional computers and supercomputers and the possibilities of efficient quantum computing. With their approach, they are attempting to accelerate quantum calculations and thus, for example, optimize construction designs according to criteria such as weight reduction, material efficiency, cost minimization and performance increase with high savings in time and cost resources.

Q-Genesys Combines Quantum Computing And Generative Learning

Advantages resulting from the physical properties of quanta, such as parallelism in processing (superposition) or, for example, entanglement, are specifically combined with solution approaches from generative learning. Generative learning attempts to achieve new, adapted solutions for optimization and simulation processes by enriching information and linking existing data appropriately.

Up to now, this has been achieved through high computational effort with classic computers known as supercomputers. However, despite the high effort involved, this did not mean sufficient acceleration in the calculation of high-dimensional probability distributions, such as those available for practical scenarios in industrial and economic processes.

Under the leadership of Siemens, the researchers compiled suitable data sets from industry and then assessed and evaluated them in practical tests. The company IQM will create quantum generative models that can be classically trained and are suitable for creating random samples.

Early Fault Tolerance And Correction Through AI Plus Quantum Computing

The experts at Fraunhofer IIS are expanding these classically generated quantum generative models in order to be able to offer the necessary strategies for early fault tolerance and correction. They primarily use artificial intelligence methods to develop and adapt error correction procedures, thereby making them as efficient as possible. In this way, the advantages offered by quantum computing can still be used on hardware that tends to operate with errors.

OTH Regensburg deals with the necessary methods for training and the adequate creation of samples that are particularly suitable for answering the problem. In this way, efficient estimation methods for the corresponding training-relevant variables or for an error-resistant sampling method can be created in the project.

Machine Authentication

Machine Operator Authentication - Why knowing "who" matters