Nanoworks

One Workflow for DFT, MD, and Machine-Learned Potentials

Configure and run computational materials simulations through a consistent, open-source Python interface.

Python-based · Open source · Reproducible workflows · Built for materials research

Computational Workflows

Nanoworks provides three focused command-line solvers while keeping calculation settings explicit and reusable.

dftsolve

DFT Workflows

Run electronic-structure and materials-property calculations through GPAW or Quantum ESPRESSO, including QE geometry optimization, spin-resolved electronic structure, projected bands, and electron-density outputs.

Explore DFT workflows →

mdsolve

Molecular Dynamics

Perform geometry optimization and molecular dynamics with classical interatomic potentials from OpenKIM.

Explore MD workflows →

mlsolve

Machine-Learned Potentials

Use MACE, CHGNet, and SevenNet for efficient structure optimization and atomistic calculations.

Explore ML workflows →

Why Nanoworks?

Consistent Inputs

Use a familiar input structure across different computational workflows.

Reproducible Workflows

Keep calculation settings explicit, readable, and reusable across materials systems.

Research-Oriented Outputs

Produce organized numerical results and publication-oriented plots for common materials analyses.

Multiple Simulation Scales

Work with first-principles, classical-potential, and machine-learned-potential calculations in one toolkit.

What Can You Calculate?

Capabilities depend on the selected solver and computational backend.

Geometry optimization Ground-state properties Band structures DOS and PDOS Projected and fat bands Spin-polarized properties Electron and spin densities Equations of state Elastic properties Phonons Optical properties Molecular dynamics ML-potential calculations

See the Examples and Usage pages for workflow-specific support and ready-to-run calculations.

A Familiar Command-Line Workflow

Provide a structure, select a solver, and keep the calculation settings in a reusable Python input file.

$ dftsolve -p 8 -g structure.cif -i input.py

The same structure-and-input pattern is used by the molecular-dynamics and machine-learned-potential solvers.

Get Started with Nanoworks

For Debian and Ubuntu systems, the automated installer prepares Nanoworks and its required scientific software:

$ curl -fsSL https://raw.githubusercontent.com/sblisesivdin/nanoworks/refs/heads/main/install_scripts/install-all-Debian-based.sh | bash

Installation Guide Run the Examples

Built on the Scientific Python Ecosystem—and More!

Nanoworks brings established scientific Python libraries together with external simulation engines such as Quantum ESPRESSO. It coordinates electronic-structure, atomistic-simulation, phonon, interatomic-potential, and machine-learning tools through consistent workflows.

ASE · GPAW · Quantum ESPRESSO · Phonopy · Elastic · OpenKIM · ASAP3 · MACE · CHGNet · SevenNet

Use Nanoworks in Your Research

If Nanoworks contributes to your work, please cite:

B. Sarikavak-Lisesivdin and S. B. Lisesivdin, “Nanoworks: A multi-scale Python-based orchestrator for materials science simulations,” Computational Condensed Matter 48, e01362 (2026).

The computational engines and libraries used in a study must also be cited. See Citing Nanoworks for the Nanoworks citation and the relevant Quantum ESPRESSO, GPAW, ASE, Phonopy, OpenKIM, Elastic, and machine-learned-potential references.

From gpaw-tools to Nanoworks

Note

Previously known as gpaw-tools. Nanoworks builds on the gpaw-tools project and extends its original ASE and GPAW workflow toward a broader computational materials platform. Read more on the About Nanoworks page.

Community

Nanoworks is open source and welcomes feedback and contributions through its GitHub repository.

You can also download the Nanoworks promotional poster for your laboratory or department.