Dresden 2026 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
TT: Fachverband Tiefe Temperaturen
TT 51: Correlated Electrons: Method Development II
TT 51.7: Talk
Wednesday, March 11, 2026, 16:45–17:00, HSZ/0101
Tensor Network Python (TeNPy) version 2: status and prospects — •Johannes Hauschild1,2, Jakob Unfried1,2, Nico Kirchner1,2, Ludwig Zweng1,2, and Frank Pollmann1,2 — 1Department of Physics, NAT school, Technical University Munich, 85748 Garching, Germany — 2Munich Center for Quantum Science and Technology (MCQST),80799 Munich, Germany
TeNPy is an established python library for the simulation of strongly correlated quantum systems with tensor networks. We report on the implementation of a new linear algebra package for the library to make it more versatile. One the one hand, it allows to switch between CPU and GPU devices, providing siginificant speedup in certain cases. On the other hand, one can toggle between different symmetry backends. Backends for none or abelian symmetries have been ported from TeNPy version 1 and available with a minimal overhead. Moreover, a new symmetry backend based on fusion trees has been implemented, which can handle general fusion categories. As such, it can be used to conserve non-abelian symmetries like SU(2), SU(n), but also to implement fermions in PEPS, or models based on anyons, e.g., the Fibonaccy chain.
Keywords: Tensor networks; Matrix product states; PEPS