ATLAS

Official Resources

• Repository/Source: Research code (referenced in literature, specific repo private/restricted)
• Key Publication: Mi et al., Comp. Phys. Comm. 200, 87 (2016)
• License: Academic/Research

Overview

ATLAS is a specialized Orbital-Free Density Functional Theory (OF-DFT) software package. Unlike Kohn-Sham DFT, ATLAS solves for the electron density directly without using orbitals, typically employing a real-space finite-difference method. This allows it to scale linearly with system size (O(N)) with a very small prefactor, enabling simulations of millions of atoms, particularly for main-group metals (like Al, Mg) where OF-DFT kinetic energy functionals are accurate.

Scientific domain: Large-scale metallurgy, Warm Consensus Matter, High-pressure physics Target user community: Method developers, Metallurgists modeling dislocations/grain boundaries

Theoretical Methods

• Orbital-Free Density Functional Theory (OF-DFT)
• Real-Space Finite-Difference method
• Wang-Teter / Thomas-Fermi-von Weizsäcker Kinetic Energy Functionals
• Local pseudopotentials (bulk-derived)
• Non-local density kernels (for kinetic energy)
• Newton-Krylov optimization algorithms

Capabilities

• System Size: Routine >100,000 atoms; Capable of millions.
• Materials: Primarily simple metals (Al, Mg, Li, Na) and their alloys.
• Output: Direct equilibrium density, total energy, forces, stress.
• Simulation: Geometry optimization, Molecular Dynamics.

Key Strengths

Linear Scaling O(N):

• By avoiding orbitals and orthogonalization, computational cost scales purely linearly with volume/atoms.
• Allows simulation of mesoscopic features (dislocations, voids) at quantum accuracy.

Robust Minimization:

• Implements advanced numerical algorithms (trust-region Newton methods) to ensure convergence of the highly non-linear OF-DFT energy functional.

Inputs & Outputs

• Inputs:
  • Cell dimensions
  • Ion positions
  • Functional choice (TF, vW, WT, etc.)
  • Local Pseudopotential parameters
• Outputs:
  • Total Energy
  • Electron Density (real space grid)
  • Forces

Interfaces & Ecosystem

• Standalone: Typically runs as a standalone solver.
• Integration: Methods often cross-pollinate with PROFESS.

Advanced Features

• Mixed Basis: Some versions explore mixed basis or adaptive grids.
• Parallelization: MPI parallelization over spatial domains.

Performance Characteristics

• Speed: Extremely fast compared to KS-DFT for large systems (orders of magnitude).
• Memory: very low memory footprint (density only, no wavefunctions).

Computational Cost

• Very Low: Can simulate thousands of atoms on a workstation.

Limitations & Known Constraints

• Accuracy: Limited by the accuracy of the Kinetic Energy Density Functional (KEDF). High accuracy only for "simple" metals (Al, Mg). Failed for Transition Metals (d-electrons) and covalent bonds (Si, C) without advanced/modern functionals.
• Pseudopotentials: Must use Local Pseudopotentials (LPS), which transfers less well than Non-Local ones.

Comparison with Other Codes

• vs PROFESS: PROFESS is the leading open-source OF-DFT code; ATLAS is a robust alternative often used for specific benchmarks or internal research.
• vs VASP: ATLAS is fundamentally different (orbital-free vs Kohn-Sham); ATLAS wins on size, VASP wins on versatility/accuracy.
• Unique strength: Specialized optimization algorithms for solving the Euler-Lagrange equation of OF-DFT.

Application Areas

• Metallurgy: Study of grain boundaries, dislocations, and stacking faults in Aluminum/Magnesium.
• Liquid Metals: Large scale MD of liquid metals.
• Warm Dense Matter: High temperature electron gas simulations.

Best Practices

• Check Material: Only apply to Al/Mg/Li unless you are testing new functionals.
• Validate KEDF: Ensure the chosen Kinetic Energy Functional reproduces KS-DFT properties for the bulk phase first.

Community and Support

• Research Based: Support is via the authors of the key papers (e.g., W. Mi, H. Wang).

Verification & Sources

Primary sources:

1. Mi, W., et al., "ATLAS: A real-space finite-difference implementation of orbital-free density functional theory", Computer Physics Communications, 200, 87-95 (2016).

Verification status: ✅ VERIFIED

• Existence: Confirmed via widely cited CPC publication.
• Availability: Research-grade distribution.