QuTeNet

Quantum computers promise a fundamental paradigm shift for extremely computationally intensive tasks. This requires the development of customised methods and powerful quantum algorithms. In the project "Quantum Tensor Networks for Quantum Simulations and Artificial Intelligence" (QuTeNet) of the DLR Institutes of Quantum Technologies, AI Security and Software Technology, an innovative architecture of quantum algorithms based on tensor networks is being researched under the leadership of the Institute of AI Security. These represent an efficient and compact method for visualising complex quantum states as a network of smaller, interconnected tensors. A particular focus of the project is on exploring this architecture for various applications, especially for solving optimisation problems, simulating quantum systems and in artificial intelligence (AI).

Motivation and challenge

Accurate simulations of the corresponding quantum systems are essential for the further development of quantum technologies and applications based on them. However, these can only be carried out accurately on classical hardware for small quantum systems (e.g. with a small number of particles). An alternative approach for accurate simulations are tensor networks, which can efficiently approximate the quantum state of the system and can be implemented on both classical and quantum hardware.

The contribution of the Institute of Quantum Technologies includes in particular the use of tensor network methods for the classical simulation and quantum simulation of quantum systems. Exemplary quantum systems are translated into a description using tensor networks and implemented on quantum hardware. In particular, applications in the field of quantum sensor technology and quantum networks are being researched. We are also investigating prospective fields of application for tensor network simulations for use in industry.

Objectives and research priorities

The team, consisting of experts in AI security, quantum technology and software technology, is pursuing several key objectives:

• Development of new concepts and methods for the implementation of tensor networks on classical hardware and quantum hardware and for the use of tensor networks for quantum simulations and AI
• Investigation of concrete use cases for the simulation of quantum systems using classical and quantum-based tensor networks and for the coupling of quantum simulations with quantum machine learning
• Evaluation of the advantages and disadvantages of quantum tensor networks compared to classical tensor networks in terms of performance, efficiency, scalability and applicability, as well as delimitation of possible areas of application for both approaches
• Exploration of prospective applications of tensor networks