Background Until scalable quantum computing hardware becomes a reality, the development of new quantum programs requires these programs to be simulated on classical computers. This simulation scales exponentially in the number of quantum bits (qubits) and linearly in the number
Background
Until scalable quantum computing hardware becomes a reality, the development of new quantum programs requires these programs to be simulated on classical computers. This simulation scales exponentially in the number of quantum bits (qubits) and linearly in the number of operations. For every additional qubit, the simulation time quadruples. A surrogate model that reduces the cost of simulation bears the potential to drastically accelerate the development of new quantum programs and algorithms.
Goal
The goal of this thesis is to investigate how well different AI architectures can take over the task of simulating a quantum program. Which architecture to choose, how to represent quantum programs, as well as the details of the training procedure are open for the student to explore.
Tasks
- Research existing literature on the use of surrogate models for quantum circuit simulation.
- Gather an overview of suitable AI architectures for the task.
- Set up a training environment and compare the previously selected architectures with one another.
Requirements
- Solid Python programming skills.
- Basic understanding of different ANN architectures and training tasks.
- Previous experience in quantum computing is helpful but not required.
Contact
Christoph Stein, Dr. Stefan Klikovits, Univ.-Prof. Dr. Manuel Wimmer