A fast and easy to use Gross-Pitaevskii equation solver.

Description

PyGPE is a CUDA-accelerated Python library for solving the Gross-Pitaevskii equations for use in simulating Bose-Einstein condensate systems.

• Documentation: https://wheelermt.github.io/pygpe-docs/

Supported features

• Scalar, two-component, spin-1, and spin-2 BEC systems.
• 1D, 2D, and 3D grid lattices.
• GPU support.
• HDF5 data saving system.
• Method for generating vortices within the system.

Requirements

• Python (3.10 and above),
• h5py (^3.6.0),
• numpy (^2.0.0),
• Matplotlib (^3.8.2)

If using a GPU:

• CUDA Toolkit (>=11.2)
• CuPy (>=10.2.0).

Installation

The simplest way to begin using PyGPE is through pip:

pip install pygpe

By default, PyGPE will use the CPU to perform calculations. However, if a CUDA-capable GPU is detected, PyGPE will automatically utilise it for drastic speed-ups in computation time.

Examples

See examples folder for various examples on the usage of the library. Below is an animation of superfluid turbulence in a scalar BEC simulated using PyGPE on a $512^2$ lattice for $N_t=200000$ time steps taking ~5 minutes to complete on an RTX 2060.