This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| CPMD | |
|---|---|
| Name | CPMD |
| Developer | IBM, Max Planck Institute, ETH Zurich, University of Cambridge |
| Released | 1980s |
| Programming language | Fortran, C |
| Operating system | Linux, Unix, macOS |
| Platform | x86, x86-64, POWER, ARM |
| Genre | Computational chemistry, Molecular dynamics, Electronic structure |
| License | Proprietary / Academic |
CPMD CPMD is a plane-wave/pseudopotential electronic structure and molecular dynamics program widely used for first-principles simulations in condensed matter physics, materials science, and theoretical chemistry. It couples Car–Parrinello molecular dynamics-style techniques with density functional theory and has been applied to problems from surface science to biomolecular solvation. The codebase and methodology intersect with research groups and institutions across Europe and North America, including collaborations that involve Max Planck Society, ETH Zurich, IBM, and the University of Cambridge.
CPMD implements ab initio molecular dynamics based on plane-wave basis sets and pseudopotentials, enabling simulations of atoms and electrons under periodic and nonperiodic boundary conditions. Users employ pseudopotentials developed by groups such as the Troullier–Martins pseudopotentials and the Vanderbilt ultrasoft pseudopotentials teams, while exchange–correlation functionals originate from formulations by Walter Kohn collaborators and the Perdew–Burke–Ernzerhof community. Typical studies integrate thermostatting methods inspired by developments from the Nosé–Hoover family and sample ensembles that connect to foundations laid by physicists including Ludwig Boltzmann and Josiah Willard Gibbs.
The theoretical foundation rests on density functional theory (DFT) as formalized by Hohenberg–Kohn and implemented via Kohn–Sham equations derived in the tradition of computational quantum mechanics. Dynamical propagation follows the Car–Parrinello approach introduced by Roberto Car and Mark Parrinello, combining fictitious electronic mass terms and ionic degrees of freedom to evolve coupled equations of motion. Electron–ion interactions are treated with norm-conserving and ultrasoft pseudopotentials developed in the laboratories of Norman Troullier and David Vanderbilt, while dispersion corrections often reference parametrizations associated with Stefan Grimme. Boundary condition and sampling strategies draw on methods from Michael Tuckerman and ensemble control schemes pioneered in the statistical mechanics community.
The implementation emphasizes plane-wave expansions and fast Fourier transforms, leveraging algorithmic advances by groups working on FFT libraries used across scientific computing, including implementations influenced by work at Argonne National Laboratory and Lawrence Livermore National Laboratory. Diagonalization routines and iterative solvers incorporate linear-algebra techniques from contributors linked to LAPACK and ScaLAPACK projects, and parallelization schemes exploit message-passing interfaces inspired by standards from the Message Passing Interface Forum. Charge-density mixing and self-consistent-field convergence build on strategies advanced by researchers at institutions like École Polytechnique Fédérale de Lausanne and University of California, Berkeley. Geometry optimization combines quasi-Newton methods traced to work by Broyden and Fletcher–Powell style algorithms that have been refined in many computational chemistry packages.
CPMD has been used to study heterogeneous catalysis on surfaces analyzed by teams at Max Planck Institute for Coal Research, electrochemical interfaces investigated in collaborations with groups at ETH Zurich, and proton transfer in aqueous systems explored by researchers associated with the University of Cambridge and University of Oxford. Other notable applications include phase transitions in oxides examined in studies linked to Paul Alivisatos-style nanomaterial research, vibrational spectroscopy predictions compared against experiments at facilities like Diamond Light Source, and biomolecular solvation problems that connect to experimental work at European Molecular Biology Laboratory. Work on battery materials and solid electrolytes often cross-references industrial research from IBM Research and national laboratories such as Oak Ridge National Laboratory.
Performance scales with plane-wave cutoff, k-point sampling, and pseudopotential complexity; benchmarks reported by computational centers including National Energy Research Scientific Computing Center show strong parallel efficiency up to node counts typical of contemporary clusters. Memory and I/O constraints reflect Fortran and C heritage shared with codes developed at Los Alamos National Laboratory. Limitations arise for systems with strong static correlation where multireference character studied by groups like Peter Reichardt or methods from the Quantum Monte Carlo community may be more appropriate, and for very large biological assemblies where classical force fields championed by developers at University of Illinois Urbana–Champaign may offer greater system size for longer timescales. Correct treatment of van der Waals interactions requires supplementary schemes influenced by research from Stefan Grimme and the Tkatchenko–Scheffler approach.
Development traces to the introduction of Car–Parrinello dynamics in the mid-1980s and subsequent collaborative software efforts among European computational chemistry groups. Key contributors include research groups at Max Planck Institute for Solid State Research, University of Cambridge, and ETH Zurich that established code frameworks integrating plane-wave DFT with molecular dynamics. The project evolved alongside parallel computing advances championed by national laboratories such as Argonne National Laboratory and software ecosystems maintained by the Open Source Initiative-adjacent scientific community. Over decades, CPMD has interfaced with pseudopotential libraries and visualization tools developed at institutions like Lawrence Berkeley National Laboratory, and its methodological lineage connects to continuing innovations in first-principles molecular dynamics pursued at universities including Yale University and University of California, Berkeley.
Category:Computational chemistry software