betaFeedback

WorksPhysics and Astronomy

Efficient Spectra and Dynamics with Tangent-Space Chebyshev Tree Tensor Networks

Abstract

We present a tangent-space Chebyshev method (TaChe), which combines tree tensor network states with Chebyshev recursion and maximum entropy spectral analysis in a tree tensor network tangent space, for quantum many-body spectra and dynamics. Implemented tensor network state calculations in the tangent space of the ground state, the core advantage of TaChe is that it can substantially reduce the computational complexity, memory usage, and numerical instability associated with tensor network state compression. Implemented on the GPU platform, it produces accurate spectra with speedups of up to 46.846.8 over the previous Chebyshev tensor network method and 13.713.7 over one-site time-dependent variational principle propagation in the reported benchmarks. Applications to acetonitrile, thermal aggregate emission, finite-temperature charge transport on a 6×66\times6 lattice, and internal conversion in aggregates of up to 100 molecules demonstrate the efficiency and versatility of TaChe in simulating complex molecular systems.

Paper details

vlnc:W_ekv32fcn
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC-BY-NC-SA-4.0
  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…