TensorBinding.jl
TensorBinding is a Julia library for tight-binding physics on matrix-product-state (MPS/MPO) representations. It targets large-scale 1D and 2D lattice models where exact diagonalisation is infeasible, combining tensor-network methods (DMRG, TDVP, KPM, TCI) with GPU acceleration.
Installation
TensorBinding is not yet registered. Install directly from the GitHub repository:
using Pkg
Pkg.add(url="https://github.com/TensorBinding/TensorBinding.jl")TensorBinding.TensorBinding — Module
TensorBinding provides tight-binding physics on MPS/MPO tensor-network representations. It combines DMRG, KPM, TDVP, and TCI with GPU acceleration for large 1D/2D lattice models.
Authors: Tiago V.C. Antão, Anouar Moustaj, Yitao Sun
Package organisation
| Section | Contents |
|---|---|
| Core | TBSystem, Hamiltonian builders, low-level utilities |
| Lattice | 2D shift operators, multilayer, twisted, flake, junction geometries |
| Solvers | KPM, time evolution, DMRG, Krylov Green's function |
| Physics | SCF, QFT, superconductivity, purification, non-Hermitian, topology, two-particle, RPA, QPI |
| GPU | CUDA-accelerated mirrors of the CPU entry points |