MaZe

Official Resources

• Homepage: https://gitlab.e-cam2020.eu/esl/MaZe
• Documentation: https://gitlab.e-cam2020.eu/esl/MaZe
• Source Repository: https://gitlab.e-cam2020.eu/esl/MaZe (E-CAM Gitlab)
• License: MIT License

Overview

MaZe is a specialized code for performing Orbital-Free Density Functional Theory Molecular Dynamics (OF-DFT-MD) using the Mass-Zero (MaZe) constrained molecular dynamics approach. It focuses on the adiabatic propagation of the electronic density, treating it as a dynamic variable with zero mass, which allows for efficient and stable time evolution of large-scale systems within the orbital-free framework.

Scientific domain: Orbital-Free DFT, Molecular Dynamics, Algorithm Development Target user community: Researchers in computational physics and extended Lagrangian dynamics

Theoretical Methods

• Orbital-Free Density Functional Theory (OF-DFT)
• Mass-Zero Constrained Dynamics
• Extended Lagrangian formulation
• Adiabatic density propagation
• Kinetic Energy Density Functionals (Thomas-Fermi, etc.)
• Local pseudopotentials

Capabilities (CRITICAL)

• OF-DFT Molecular Dynamics
• Constant energy (NVE) ensembles
• Efficient density propagation
• Large-scale metallic systems
• Integration with standard MD workflows

Key Strengths

Mass-Zero Algorithm:

• Avoids minimization of energy at every time step
• Propagates density constraints
• Maintains Born-Oppenheimer surface adherence
• Improved computational efficiency for dynamics

Implementation:

• High-Performance Computing (HPC) ready
• C/C++ implementation
• MPI parallelization

Inputs & Outputs

• Input:
  • Simulation parameters
  • Ionic configurations
  • Pseudopotentials
• Output:
  • Trajectories
  • Conserved quantities (Energy)
  • Density snapshots

Interfaces & Ecosystem

• E-CAM: Part of the E-CAM software library for HPC
• Libraries: Uses standard numerical libraries (FFTW, BLAS)

Advanced Features

• Curvilinear Coordinates: Generalization of MaZe to curvilinear constraints
• Stability: Enhanced stability over standard Born-Oppenheimer MD in some regimes

Performance Characteristics

• Speed: Efficient propagation step
• Scaling: O(N) due to Orbital-Free nature
• Accuracy: Consistent with OF-DFT models

Computational Cost

• Complexity: Linear with system size
• Overhead: Low overhead per timestep

Limitations & Known Constraints

• Method Scope: Specialized for MaZe algorithm testing and usage
• Functionals: Limited to available OF-DFT functionals
• Documentation: Developer-focused

Comparison with Other Codes

• vs PROFESS: PROFESS is a general purpose OF-DFT suite; MaZe focuses on the specific dynamics algorithm
• vs CP2K: CP2K uses orbital-based minimization (OT); MaZe uses orbital-free constraints
• Unique strength: Mass-Zero constrained dynamics implementation for OF-DFT

Application Areas

• Liquid Metal Dynamics: Efficient sampling of phase space
• Algorithm Research: Testing extended Lagrangian methods
• Large Scale MD: Thousand-atom simulations

Verification & Sources

Primary sources:

1. Repository: https://gitlab.e-cam2020.eu/esl/MaZe
2. References in E-CAM documentation
3. A. M. P. et al., J. Chem. Phys. (Related methodology papers)

Confidence: VERIFIED

• Status: Available Open Source
• Project: E-CAM Center of Excellence