| GAMESS (US) | |
|---|---|
| Name | GAMESS (US) |
| Developer | Mark S. Gordon group; contributors at Iowa State University, University of California, Berkeley, University of Washington, and national laboratories |
| Released | 1981 (origins) |
| Latest release | ongoing development (community-maintained) |
| Programming language | Fortran |
| Operating system | Unix, Linux, Microsoft Windows |
| Genre | Computational chemistry / Quantum chemistry |
| License | Academic source-available (historically free for academics); varied contributor licenses |
GAMESS (US)
GAMESS (US) is a general-purpose electronic structure program for performing ab initio molecular quantum chemistry calculations. It is one of the longstanding packages alongside Gaussian, Molpro, NWChem, and Q-Chem that enabled the modern practice of computational quantum chemistry and contributed to advances in Quantum Physics through accurate many-electron simulations. Originating from code developed in the late 1970s and early 1980s, GAMESS (US) grew from collaborative efforts at institutions including Iowa State University, the University of California, Berkeley, and several United States Department of Energy laboratories to support methods such as Hartree–Fock, MP2, and coupled cluster techniques. The project evolved in parallel to the separate GAMESS (UK) fork, reflecting different community and development models.
GAMESS (US) implements a broad suite of quantum chemical methods used in theoretical chemistry and computational physics. Core algorithms include self-consistent field (SCF) solutions of the Hartree–Fock equations, density functional theory (DFT) with multiple exchange–correlation functionals, and correlated wavefunction approaches such as CI, Møller–Plesset perturbation theory, and coupled cluster (CCSD, CCSD(T)). The package contains implementations of analytic gradients and Hessians that enable geometry optimization and vibrational frequency analysis, and supports excited-state techniques including CIS, equation-of-motion coupled cluster (EOM-CC), and linear-response TD-DFT. For efficient evaluation of two-electron integrals it uses integral-direct algorithms and density fitting / resolution-of-the-identity approaches, connecting to numerical linear algebra methods used in high-performance computing such as BLAS and LAPACK.
GAMESS (US) is predominantly written in Fortran and organized as modular code blocks that handle integral generation, SCF iterations, post-Hartree–Fock solvers, and I/O. It supports parallel execution via distributed-memory paradigms and has been ported to supercomputing platforms at facilities like the Argonne National Laboratory and Oak Ridge National Laboratory. The input language is plaintext with a keyword-oriented structure familiar to computational chemists. GAMESS integrates with basis set libraries such as the Basis Set Exchange standards and supports effective core potentials and relativistic corrections for heavy-element simulations (e.g., scalar relativistic approximations). Interoperability options include interfaces for visualization and workflow tools used in materials and chemical informatics.
Researchers use GAMESS (US) for studies ranging from small-molecule spectroscopy to reaction mechanisms, noncovalent interactions, and materials-relevant clusters. In quantum dynamics and spectroscopy, GAMESS supplies potential energy surfaces, transition moments, and vibrational analyses for input to dynamical simulations. It has been applied in investigations of catalysis, atmospheric chemistry, and biomolecular active sites in collaboration with experimental groups at institutions like Lawrence Berkeley National Laboratory and universities such as University of Illinois Urbana–Champaign. GAMESS outputs support study of electron correlation effects that are central to many-body physics, and its methodologies inform model Hamiltonians in condensed-matter contexts and molecular-scale components in quantum information research.
Performance studies of GAMESS (US) compare its accuracy and computational cost against packages such as Gaussian, Molpro, and ORCA. Benchmark suites use standardized test sets for thermochemistry, noncovalent interactions (e.g., S22), and barrier heights, evaluating methods like CCSD(T) as the “gold standard.” Validation includes reproduction of spectroscopic constants, potential energy curves, and comparison to high-level quantum Monte Carlo or experimental data from institutions like the NIST. Parallel-scaling benchmarks have guided optimizations for multi-core clusters and leadership-class machines, emphasizing equitable access by documenting performance on modest university clusters as well as on systems at XSEDE and other computing infrastructures.
GAMESS (US) historically adopted an academic source-available model permitting free use by academics and collaborators while maintaining controlled redistribution. The development community comprises university groups, national laboratories, and independent contributors; governance is informal and collaborative. Training materials, user forums, and workshops at conferences such as the American Chemical Society meetings and the Gordon Research Conferences have supported adoption. Efforts to improve accessibility include cross-platform builds, containerization with technologies like Docker and Singularity, and documentation to lower barriers for researchers at resource-limited institutions.
GAMESS (US) plays a role in science education by enabling hands-on computational projects in undergraduate and graduate curricula, bridging theory and experiment for students at diverse institutions. Because computational resources and software licenses can be inequitable, the GAMESS (US) community has been part of broader conversations about open science, reproducibility, and equitable access to computational infrastructure. Workshops and mentoring programs aim to expand participation by historically underrepresented groups in STEM, partnering with initiatives at public universities and national labs to democratize training in electronic structure methods. Continued emphasis on transparent algorithms, reproducible input/output practices, and interoperable data formats supports socially responsible research in quantum chemistry and physics.
Category:Computational chemistry software Category:Quantum chemistry