Installation

The quatrex package is an ab initio quantum transport simulator written in Python. We leverages Python's core CPU and GPU array frameworks (numpy and cupy) as well as associated frameworks and libraries (scipy, mpi4py, numba). The quatrex codebase is designed to be extensible, portable, and highly performant. It shows excellent scaling and sustained exascale performance on different supercomputers.1

Here we provide instructions for setting up a basic environment for running quantum transport simulations with quatrex.

Note

We currently only provide instructions for installing quatrex from source and with conda.

First, clone the repository

git clone https://github.com/quatrex/quatrex.git
cd quatrex

We provide several example configurations, input files, and reference outputs that we use for testing and development. You can find them in the examples/ directory.

We track this data using Git LFS, so make sure to install Git LFS and pull the files after cloning the repository:

git lfs install
git lfs pull

You can then create a conda environment from the provided environment-dev.yml file

conda env create -f environment-dev.yml
conda activate quatrex

This basic environment includes numpy and scipy as dependencies. It does not include mpi4py and cupy by default. The reason for this is that you may want to leverage your system's MPI and GPU backend. This is especially the case in HPC environments.

To leverage your system's backends, you will have to install mpi4py and cupy from source (e.g. via PyPI):

pip install mpi4py
pip install cupy  # (for GPU support)

If you do not care about using your system's backend, you can install mpi4py and cupy from the conda-forge channel:

conda install -c conda-forge mpi4py mpich
conda install -c conda-forge cupy cuda-version=XX.X  # (for GPU support)

Finally, you can install quatrex from source:

pip install .

  1. N. Vetsch, et al., Ab-initio Quantum Transport with the GW Approximation, 42,240 Atoms, and Sustained Exascale Performance, arXiv:2508.19138 (2025). ↩