**Type:**Master thesis**State:**running**Projects:**TAQO-PAM**Supervisor:**Wolfgang Mauerer, Ralf Ramsauer**Student:**Andrej Utz

The current state of QC technology is far from ideal: isolating multiple qubits from inﬂuence, such as quantum noise by eliminating noise factors or maintaining their coherence, is a challenging task. [1] Additionally the number of qubits of a QC is the one of the limiting factor for use in real-world applications. A well known and studied approach is the Quantum Approximate Optimisation Algorithm (QAOA) [2], which utilises hybrid quantum-classical systems. A CPU manages, in tandem with a GPU, the optimisation process to approximate the optimal solution. By splitting up the computation into separate tasks and applying a ﬁtting solver for each of them, QAOA incorporates the best aspects from both worlds, classical and quantum. However, while the result may be computed faster, we cannot expect the execution time to be anticipated exactly in all cases due to multiple inﬂuences. Wintersperger et al. formalised [3] these as:

The authors also showed the optimisation possibilities, especially from a QPU design perspective, by analysing the qubit coupling density. We will focus on the classical HPC part: 𝑡_comm and 𝑡_opt.

The objective of this work is to reﬁne the presented runtime model and to ﬁnd more optimisation possibilities in the classic component using statistical model. We want to provide a break-even point for real world applications.

[1] J. Preskill, “Quantum Computing in the NISQ era and beyond”, Quantum, p. 79, Aug. 2018, doi: 10.22331/q-2018-08-06-79.

[2] E. Farhi and A. W. Harrow, “Quantum Supremacy through the Quantum Approximate Optimiza- tion Algorithm”. 2019.

[3] K. Wintersperger, H. Saﬁ, and W. Mauerer, “QPU-System Co-design for Quantum HPC Accelera- tors”, M. Schulz, C. Trinitis, N. Papadopoulou, and T. Pionteck, Eds., Cham: Springer International Publishing, 2022, pp. 100–114.