MuST

Official Resources

• Homepage: https://github.com/mstsuite/MuST
• Documentation: https://must.readthedocs.io/
• Source Repository: https://github.com/mstsuite/MuST
• License: BSD 3-Clause

Overview

MuST is an open-source ab initio electronic structure suite based on Multiple Scattering Theory (MST). It integrates the Korringa-Kohn-Rostoker (KKR) Green function method and the Locally Self-consistent Multiple Scattering (LSMS) method. It is unique in its ability to handle disordered materials (via CPA) and scale to petascale/exascale computing systems for massive all-electron calculations.

Scientific domain: Disordered materials, alloys, magnetic systems, quantum phase transitions Target user community: Researchers in metallurgy, disordered systems, and high-performance computing

Theoretical Methods

• Density Functional Theory (DFT)
• Multiple Scattering Theory (MST) / KKR
• Coherent Potential Approximation (CPA) for random alloys
• Locally Self-consistent Multiple Scattering (LSMS)
• Kubo-Greenwood formula for conductivity
• Landau-Lifshitz-Gilbert (LLG) dynamics for spins
• Full-potential and Muffin-tin approximations
• Relativistic effects (Scalar and Fully Relativistic)

Capabilities

• Electronic structure of ordered and disordered solids
• Linear-scaling (O(N)) calculations for tens of thousands of atoms
• Electrical conductivity in random alloys
• Spin dynamics and thermodynamics
• First-principles calculation of critical temperatures
• Defect states and impurities

Key Strengths

Disordered Materials

• KKR-CPA method efficiently averages over disorder without supercells
• Direct calculation of alloy properties

Massive Scaling (LSMS)

• O(N) scaling allows simulation of extremely large systems
• Designed for top-tier supercomputers (Exascale ready)
• GPU acceleration

Inputs & Outputs

• Input: Fortran-namelist style inputs, atom positions, potential files
• Output: DOS, Band structure (KKR), conductivity, magnetic moments

Interfaces & Ecosystem

• Input/Output:
  • Fortran namelist input format.
  • HDF5 support for large data handling.
  • Integration with multiple scattering theory analysis tools.
• Libraries:
  • Uses LAPACK/BLAS, MPI, and OpenMP.

Advanced Features

• KKR-CPA:
  • Coherent Potential Approximation for rigorous treatment of random alloys.
  • Handles chemical disorder without large supercells.
• Kubo-Greenwood Transport:
  • First-principles calculation of electrical conductivity.
  • Residual resistivity in alloys.
• Landau-Lifshitz-Gilbert (LLG):
  • Spin dynamics simulations for magnetic systems.
  • Thermodynamic properties of magnets.

Community and Support

• Documentation: Comprehensive guides at https://must.readthedocs.io/
• Development: Managed by Oak Ridge National Lab (ORNL) and collaborators.
• Support: GitHub issues and documentation tutorials.

Computational Cost

• LSMS: Linear scaling O(N), highly efficient for very large supercells.
• KKR: Efficient for periodic unit cells and alloys using Green's functions.

Verification & Sources

Primary sources:

1. GitHub Repository: https://github.com/mstsuite/MuST
2. Documentation: https://must.readthedocs.io/
3. "MuST: An open source package for ab initio electronic structure calculations in diverse computing environments"

Confidence: VERIFIED Status: Open Source, Active (ORNL/NSF supported)