GPU-accelerated simulations of quantum annealing and the quantum approximate optimization algorithm

We study large-scale applications using a GPU-accelerated version of the massively parallel J\"ulich universal quantum computer simulator (JUQCS--G). First, we benchmark JUWELS Booster, GPU cluster with 3744 NVIDIA A100 Tensor Core GPUs. Then, use JUQCS--G to relation between annealing (QA) and approximate optimization algorithm (QAOA). find that very coarsely discretized QA, termed (AQA), performs surprisingly well in comparison QAOA. It can either be used initialize QAOA, or avoid costly procedure altogether. Furthermore, scaling success probability when AQA for problems 30 40 qubits. case largest discretization error scales most favorably, surpassing best result obtained from