Year: 2023
Type: dissertation
Abstract: We propose and benchmark a matrix product operator (MPO) based time evolution which uses a variational QR decomposition based truncation scheme instead of a singular value decomposition (SVD). This improves the scaling of the time evolution algorithm with respect to the local Hilbert space dimension d from d3 to d2. Additionally, we demonstrate that the proposed algorithm runs efficiently on GPUs in contrast to the established SVD based truncation method. This results in an additional, hardware-dependent speedup.
Language: English
Topic: Quantum many-body systems
Subfield: Atomic and Molecular Physics, and Optics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7066069147