QuantumATK X-2025.06: Feature Overview

QuantumATK X-2025.06 is a leading platform for atomic-scale modeling, integrating density functional theory (DFT), semi-empirical tight binding, classical force fields, and machine-learned force fields (ML-FFs). It accelerates research and development in semiconductors, materials science, and nanotechnology by enabling large-scale, realistic simulations with enhanced performance.

Key Features of QuantumATK X-2025.06

  1. Advanced Simulation Engines:
    • Supports DFT with linear combination of atomic orbitals (LCAO) and plane-wave basis sets, semi-empirical models, and classical potentials for versatile material simulations.
    • Non-Equilibrium Green’s Function (NEGF) module for simulating nanoscale devices and interfaces, including support for non-zero bias, electrostatic gates, and dielectrics.
    • Enables modeling of complex systems like high-k metal gate stacks and 2D material-based field-effect transistors (FETs).
  2. Performance Improvements:
    • Achieves 10x+ speedup in critical calculations (e.g., self-consistent field (SCF), bandstructure, PDOS, PLDOS, transmission spectrum) for bulk and NEGF device simulations.
    • Up to 5x faster PDOS, FatBandstructure, and Magnetic Anisotropy Energy (MAE) analysis, with speed increasing for larger projection numbers.
    • Supports multi-node and multi-GPU parallelization with linear acceleration, reducing runtime for large systems (e.g., 5,000 atoms with DFT or 30,000 atoms with semi-empirical NEGF) from days to hours.
    • 2x speedup for SCF and geometry optimization with MetaGGA, GGA, and Hubbard U for small to medium systems (a few hundred atoms).
  3. Machine-Learned Force Fields (ML-FFs):
    • Enables simulations of large-scale systems (10,000–100,000+ atoms) and long-timescale dynamics when conventional force fields are unavailable or ab initio methods are too costly.
    • Pre-trained ML-FFs for materials like Si, SiO2, HfO2, TiN, TiSi, TiNAlO, and interfaces (e.g., TiN|AlO, Si|SiO2, HfO2|TiN).
    • Supports hybrid molecular dynamics (MD)/force-bias Monte Carlo (FBMC) for enhanced equilibration in deposition simulations (e.g., HfCl4 on HfO2 surfaces).
  4. Enhanced Analysis Tools:
    • Automated workflows for dynamical matrix (D) and Hamiltonian derivatives (dH/dR) with Wigner-Seitz scheme for faster calculations and improved accuracy in large systems.
    • Tetrahedron integration for mobility and resistivity calculations of complex Fermi surfaces, with approximate methods for phonon-limited resistivity (constant mean-free path for nanostructures, constant relaxation time for bulk).
    • Special Thermal Displacement (STD) method for electron-phonon coupling, incorporating phonon mode weighting at various temperatures.
  5. User Interface and Workflow:
    • NanoLab GUI provides an intuitive interface for model setup, simulation, and visualization, reducing focus on technical details.
    • Python-based scripting via atkpython executable for advanced customization and automation.
    • Supports flexible boundary conditions (Dirichlet, von Neumann, periodic) and k-space symmetry utilization for reduced computational time.
    • Built-in framework for defect migration paths, transition states, and reaction energies using Nudged Elastic Band (NEB) method.
  6. Applications:
    • Used in semiconductor R&D, battery design, catalysis, polymers, and nanotechnology for simulating material properties, device performance, and atomic-scale interactions.
    • Supports industries like electronics, energy, and materials science with tools for optimizing material options.

image 34