GAMESS

GAMESS
Name	GAMESS
Developer	Samuel N. Waldeck; Mark Gordon; Jan Martin Paldus
Released	1970s
Programming language	Fortran (programming language); MPI (message passing interface)
Operating system	Linux; Microsoft Windows; macOS
Genre	Computational chemistry; Electronic structure theory
License	Proprietary; Free academic distribution

GAMESS GAMESS is a quantum chemistry package for electronic structure calculations used in computational and theoretical chemistry, molecular physics, and materials science. It provides methods for molecular energy, structure, spectroscopy, and dynamics predictions and is developed and maintained by research groups associated with national laboratories and universities. The program is widely used in academic, industrial, and governmental laboratories for research connected to spectroscopy, catalysis, reactive intermediates, and materials modeling.

History

GAMESS originated from work in the 1970s at research centers including University of California, Berkeley, Iowa State University, and collaborations with Lawrence Berkeley National Laboratory and Argonne National Laboratory, reflecting influences from figures associated with John Pople-era developments, Pulay-related optimization methods, and contemporaneous packages such as Gaussian (software), MOLPRO, NWChem, Psi4, and ORCA (computational chemistry package). Over time contributions came from scientists affiliated with University of Tennessee, Iowa State University, Australian National University, Rutherford Appleton Laboratory, and national research initiatives funded by agencies like National Science Foundation (United States) and Department of Energy (United States). Key developments in parallelization and algorithmic innovation paralleled advances at institutions such as Sandia National Laboratories, Lawrence Livermore National Laboratory, and collaborations engaging researchers from Rutgers University and University of Cambridge.

Features and Capabilities

The package implements self-consistent field frameworks rooted in methods similar to those advanced by Walter Kohn and Lev Landau-inspired many-body approaches, offering Hartree–Fock, density functional theory approximations comparable to functionals proposed by John Perdew, Axel Becke, and Lee Yang Parr, and post-Hartree–Fock correlation methods in line with formalisms used by Kurt Burke and Christopher C. J. Roothaan. GAMESS supports analytic gradients, vibrational frequency analysis, excited-state properties consistent with theories developed by Marius Schütz and Angel Rubio, and solvation models related to concepts advanced by researchers at University of California, Santa Barbara and ETH Zurich. Interfaces and interoperability are maintained with visualization and workflow tools developed at centers such as Argonne National Laboratory, Brookhaven National Laboratory, and software ecosystems including Jupyter Notebook, VMD (software), and Avogadro (software).

Theoretical Methods and Algorithms

The codebase includes configuration interaction and coupled cluster approaches reflecting formulations from Ira N. Levine-style textbooks, similarity-transformed theories akin to developments by Robert J. Bartlett and Gustav K. Schenter, and multiconfigurational self-consistent field machinery influenced by work at Max Planck Institute for Coal Research and École Normale Supérieure. Perturbation theories in GAMESS trace conceptual lineage to methods like Møller–Plesset perturbation originating from Christian Møller and Møller–Plesset, while response theories and property evaluation implement frameworks comparable to techniques used by groups at University of Oxford and Columbia University. Integrals and basis set treatments employ conventions established by Frank Jensen and basis set families associated with researchers at Ditchfield-era groups and later expansions from Pople-influenced basis development.

Software Architecture and Implementation

GAMESS is implemented primarily in Fortran (programming language) with parallelization using MPI (message passing interface) paradigms developed alongside standards influenced by Argonne National Laboratory and Lawrence Livermore National Laboratory practices. The modular code structure reflects software-engineering practices promoted by projects at Sandia National Laboratories and Los Alamos National Laboratory, enabling integration with linear algebra libraries from Netlib-affiliated projects and vendor packages like Intel Math Kernel Library and OpenBLAS. Build and portability tools draw on routines and conventions from GNU Compiler Collection ecosystems and continuous-integration practices akin to efforts at CERN and Software Carpentry-style training initiatives.

Performance and Benchmarks

Performance optimizations target distributed-memory clusters and many-core architectures, with benchmark comparisons performed on systems from vendors represented at conferences such as SC (conference) and ACM-sponsored events. Scaling studies reference hardware from Cray Research, IBM, and Intel and compare timings with other packages like NWChem, ORCA (computational chemistry package), Gaussian (software), and Q-Chem. Benchmarks commonly involve standard test sets from communities associated with Gaussian (software)-era validation suites, and performance tuning often leverages techniques developed at Lawrence Berkeley National Laboratory and National Center for Supercomputing Applications.

Applications and Use Cases

GAMESS is applied to electronic spectra problems similar to work at Max Planck Institute for Chemical Physics of Solids and reaction-path studies in line with research from California Institute of Technology, Massachusetts Institute of Technology, and Stanford University. It supports studies of catalytic cycles investigated at Fritz Haber Institute and ETH Zurich, photochemistry research parallel to groups at Imperial College London and University of Cambridge, and materials-related simulations connecting to projects at Argonne National Laboratory and Oak Ridge National Laboratory. Industrial applications draw on examples from Pfizer, BASF, and Dow Chemical Company-style research programs in molecular design and property prediction.

Licensing and Distribution

Distribution models evolved through academic and national-laboratory licensing practices similar to arrangements at University of California-affiliated software and other community codes originating from Department of Energy (United States) laboratories. Binary builds and source distributions are maintained for platforms including Linux, Microsoft Windows, and macOS, with user support resembling community practices at Stack Overflow and collaborative development workflows influenced by GitHub-era norms. Academic licenses and site-use agreements parallel policies at institutions such as University of Texas and Yale University for research computing.

Category:Computational chemistry software