Qbox (TDDFT)

Official Resources

• Homepage: http://qboxcode.org/
• Documentation: http://qboxcode.org/doc/html/
• Source Repository: https://github.com/qboxcode/qbox-public
• License: GNU General Public License v2.0

Overview

Qbox is a scalable parallel implementation of first-principles molecular dynamics based on the plane-wave, pseudopotential formalism. Developed by François Gygi at UC Davis, Qbox is specifically designed for exceptional parallel scalability on high-performance computing systems, with demonstrated efficiency on tens of thousands of processors. It excels at large-scale ab initio molecular dynamics simulations of complex systems, particularly for studying materials under extreme conditions.

Scientific domain: First-principles MD, plane-wave DFT, extreme scalability, HPC
Target user community: Materials scientists, HPC researchers, extreme conditions, large-scale MD

Theoretical Methods

• Kohn-Sham DFT (LDA, GGA, hybrid functionals)
• Plane-wave basis with pseudopotentials
• Norm-conserving pseudopotentials
• Born-Oppenheimer molecular dynamics
• Car-Parrinello-like dynamics
• Variable-cell dynamics
• Path integral molecular dynamics (PIMD)
• van der Waals corrections (vdW-DF)
• DFT+U for correlated systems
• Meta-GGA functionals
• Exact exchange (HFX) for hybrids
• Spin-orbit coupling
• Non-collinear magnetism
• Ultrasoft pseudopotentials (limited)

Capabilities (CRITICAL)

• Ground-state electronic structure calculations
• Born-Oppenheimer molecular dynamics
• First-principles molecular dynamics at scale
• Geometry optimization
• Real-time time-dependent DFT
• Optical absorption spectra via RT-TDDFT
• Electronic stopping power calculations
• Forces and stress tensors
• Parallel execution on 1000s of processors
• XML-based checkpointing and I/O
• Wave function extrapolation for MD
• Variable cell dynamics
• Thermostats and barostats
• Optimized for large systems (1000+ atoms)

Sources: Official Qbox documentation, cited in 6/7 source lists

Inputs & Outputs

• Input formats:

  • XML input files (native format)
  • Interactive command-line interface
  • Coordinate files (XYZ)
  • Pseudopotential files (XML format)

• Output data types:

  • XML output with all computed properties
  • Sample files (snapshots during MD)
  • Wavefunction checkpoints
  • Trajectory files
  • Energy and force outputs

Interfaces & Ecosystem

• Framework integrations:

  • Can be interfaced with workflow systems
  • XML I/O allows custom parsing

• HPC optimization:

  • Massively parallel (MPI)
  • Optimized for Cray, IBM, Intel architectures
  • ScaLAPACK for linear algebra
  • FFTW for fast Fourier transforms

• Analysis:

  • qbox_xyz - trajectory extraction
  • Custom XML parsers
  • Standard molecular dynamics analysis tools

Limitations & Known Constraints

• Specialized focus: Optimized for MD; fewer features than general DFT codes
• Input format: XML-based, requires learning curve
• Pseudopotentials: Limited to specific formats (norm-conserving primarily)
• Post-processing: Fewer established analysis tools compared to VASP/QE
• Documentation: Good but less extensive than major codes
• Community: Smaller user base
• Hybrid functionals: Limited implementation
• Features: Focuses on MD; fewer property calculations than general codes
• Basis sets: Plane-wave only
• Platform support: Primarily for HPC systems; not optimized for desktop use

Computational Cost

• Scalability: Scaling to 100,000+ cores demonstrated (Blue Gene/Q).
• Efficiency: Optimized for large-scale AIMD (1000+ atoms).
• Overhead: Minimum system size recommended is ~50 atoms due to parallel overhead structure.

Comparison with Other Codes

• vs Quantum ESPRESSO: Qbox is designed specifically for "First Principles MD at scale". QE is a general purpose swiss-army knife. Qbox is simpler but faster for its specific niche (large AIMD).
• vs CPMD: Qbox uses Born-Oppenheimer MD (mostly), CPMD uses Car-Parrinello. Qbox scales better on modern massive supercomputers.

Best Practices

• Input: Learn the XML-based input or use the interactive command line for steering simulations.
• HPC: Use the row_m, col_m matrix distributions to map the calculation to the specific torus/interconnect topology of your supercomputer.
• Restart: Qbox's XML restart files are robust; use them frequently to checkpoint long MD runs.

Community and Support

• Mailing List: qbox-users list available.
• Documentation: Online HTML manual is the primary resource.

Verification & Sources

Primary sources:

1. Official website: http://qboxcode.org/
2. Documentation: http://qboxcode.org/doc/
3. GitHub repository: https://github.com/qboxcode/qbox-public
4. F. Gygi, IBM J. Res. Dev. 52, 137 (2008) - Qbox architecture
5. E. Draeger et al., J. Parallel Distrib. Comput. 106, 205 (2017) - Scaling study

Secondary sources:

1. Qbox tutorials and examples
2. Published AIMD applications using Qbox
3. HPC benchmarking studies
4. Confirmed in 6/7 source lists (claude, g, gr, k, m, q)

Confidence: CONFIRMED - Appears in 6 of 7 independent source lists

Verification status: ✅ VERIFIED

• Official homepage: ACCESSIBLE
• Documentation: ACCESSIBLE
• Source code: OPEN (GitHub)
• Community support: Active (mailing list, development team)
• Academic citations: >200 (main papers)
• HPC pedigree: Developed at LLNL, optimized for supercomputers