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.
| CASTEP | |
|---|---|
| Name | CASTEP |
| Developer | Lancaster University; STFC; Intel Corporation |
| Released | 1990s |
| Latest release | 2020s |
| Programming language | Fortran (programming language) |
| Operating system | Linux, Unix, macOS |
| License | Proprietary / Academic |
CASTEP CASTEP is a plane-wave pseudopotential code for first-principles materials modelling based on density functional theory and related methods. Widely used in computational materials science, computational chemistry, and condensed matter physics, it serves researchers at institutions such as University of Cambridge, Imperial College London, University of Oxford, Argonne National Laboratory, and industrial partners like Siemens and IBM. Development intersects initiatives at STFC Rutherford Appleton Laboratory, collaborations with European Union projects, and contributions from research groups at University of Bath, University of Warwick, and University of Bristol.
CASTEP performs total-energy, force, stress, electronic-structure, and response-property calculations for solids, surfaces, nanostructures, and molecules using plane-wave basis sets and pseudopotentials. It contrasts with localized-basis codes developed at Bell Labs, methods from Los Alamos National Laboratory, and all-electron approaches exemplified by Wien2k and ELK (software). Users run CASTEP on high-performance computing platforms such as ARCHER UK National Supercomputing Service, NCSA, Oak Ridge National Laboratory, Lawrence Berkeley National Laboratory, and cloud providers affiliated with Amazon Web Services and Microsoft Azure.
Built on Kohn–Sham equations, CASTEP implements density functional perturbation theory for phonons and linear response, time-dependent extensions related to time-dependent density functional theory, and many-body corrections including GW approximation and hybrid functionals akin to those used by Perdew–Burke–Ernzerhof and Heyd–Scuseria–Ernzerhof. Pseudopotentials used in CASTEP follow schemes from Troullier–Martins pseudopotentials, Vanderbilt ultrasoft pseudopotentials, and projector augmented-wave approaches inspired by Peter E. Blöchl. Exchange-correlation functionals implemented include parametrizations linked to John P. Perdew, Walter Kohn, and the Local Density Approximation tradition originating near Bell Labs and Princeton University.
CASTEP offers geometry optimization, molecular dynamics, phonon calculations, dielectric tensors, Raman intensities, NMR chemical shifts, and optical spectra. It supports point-defect modelling, surface slab calculations, and transition-state searches comparable to techniques used in Gaussian (software), VASP, and Quantum ESPRESSO. Post-processing tools interface with visualization packages like VESTA, XCrySDen, Avogadro (software), and analysis suites from Materials Project and The OpenKIM Project. CASTEP’s workflows integrate with materials-data infrastructures such as NOMAD Laboratory, AFLOW, Citrine Informatics, Harvard Clean Energy Project, and MATLAB toolboxes used at Massachusetts Institute of Technology.
The codebase is primarily in Fortran (programming language) and parallelized with Message Passing Interface for distributed-memory execution on petascale systems. Fast Fourier transforms use libraries from FFTW, Intel MKL, and vendor-optimized kernels from NVIDIA for GPU offload in collaboration with Oak Ridge National Laboratory initiatives. Electronic minimization employs algorithms related to Davidson algorithm, conjugate gradient method popularized at Los Alamos National Laboratory, and Pulay-style mixing strategies traced to Peter Pulay (chemist). Phonon and response routines draw on formulations from Baroni, S., Giannozzi, P., and researchers affiliated with SISSA and CNR.
Researchers apply CASTEP to study semiconductor band structures in materials investigated at Bell Labs and Intel Corporation, battery electrode materials researched at Argonne National Laboratory and Oak Ridge National Laboratory, photovoltaic absorbers examined by groups at California Institute of Technology and University of California, Berkeley, and heterogeneous catalysts studied at Max Planck Society institutes and ETH Zurich. It appears in publications on superconductivity related to experiments at CERN, high-pressure phases explored at Diamond Light Source and European Synchrotron Radiation Facility, and surface science linked to IBM Research and Sandia National Laboratories. Industrial deployments include design problems at Rolls-Royce and BP.
CASTEP is distributed under academic licensing terms from the commercial entity associated with Cambridge Enterprise and managed development involving contributors at University of Cambridge, Lancaster University, and STFC Rutherford Appleton Laboratory. Development roadmaps have been influenced by funding from UK Research and Innovation, European framework programs like Horizon 2020, and collaborations with hardware partners including Intel Corporation and NVIDIA. Training and dissemination occur via courses at University of Cambridge, summer schools hosted by Psi-k, and workshops at conferences organized by Materials Research Society and American Physical Society.
Benchmark studies compare CASTEP to plane-wave codes such as VASP, Quantum ESPRESSO, and ABINIT across test suites from Materials Project and Quantum Mobile datasets. Scalability analyses report efficient strong and weak scaling on systems like ARCHER, Blue Waters, and Summit (supercomputer), with performance tuned against vendor libraries produced by Cray Inc., Intel, and NVIDIA. Accuracy benchmarks for cohesive energies, band gaps, and phonon dispersions reference experimental standards from NIST and collaborative databases managed by ICSD and Pearson's Crystal Data.
Category:Computational physics software