NWChem: Delivering High-Performance Computational Chemistry
NWChem aims to provide its users with computational chemistry tools that are scalable both in their ability to treat large scientific computational chemistry problems efficiently, and in their use of available parallel computing resources from high-performance parallel supercomputers to conventional workstation clusters.
NWChem software can handle
- Biomolecules, nanostructures, and solid-state
- From quantum to classical, and all combinations
- Ground and excited-states
- Gaussian basis functions or plane-waves
- Scaling from one to thousands of processors
- Properties and relativistic effects
NWChem is actively developed by a consortium of developers and maintained by the EMSL located at the Pacific Northwest National Laboratory (PNNL) in Washington State. Researchers interested in contributing to NWChem should review the Developers page. The code is distributed as open-source under the terms of the Educational Community License version 2.0 (ECL 2.0).
The current version of NWChem is version 6.3 can be downloaded here.
The NWChem development strategy is focused on providing new and essential scientific capabilities to its users in the areas of kinetics and dynamics of chemical transformations, chemistry at interfaces and in the condensed phase, and enabling innovative and integrated research at EMSL. At the same time continued development is needed to enable NWChem to effectively utilize architectures of tens of petaflops and beyond.
Ongoing Projects and Future Directions
Density functional theory (DFT), time-dependent DFT (TD-DFT) and properties
- Discrete interaction model/quantum mechanical method (DIM/QM) for describing the response properties of molecules adsorbed on metal nanoparticles. Developers: Justin Moore, Lasse Jensen (Penn State University).
- Development of exact two-component relativistic theory and calculations of magnetic response parameters. Developers: Jochen Autschbach (SUNY Buffalo).
- Development of self-consistent state-specific vertical excitation model (VEM) for electronic excitation in solution and solvatochromatic shifts in liquid-phase absorption spectra. Developers: Alek Marenich, Chris Cramer, Don Truhlar (University of Minnesota), Niri Govind (PNNL).
- Generalization of real-time TDDFT to include spin-orbit effects . Developers: Niri Govind (PNNL), Ken Lopata (LSU).
- Developing infrastructure for incorporating new density functionals and higher order derivatives thereof. The idea is to extend the density functionals in NWChem to support higher order partial derivatives to support new functionality. At the same this is a good opportunity to build the infrastructure needed to incorporate new density functionals and their higher order derivatives. The aim is to use open source tools as much as possible to make it easy for anyone to do this. Developers: Huub van Dam (PNNL).
- Exchange-hole dipole moment (XDM) method. Developers: Alberto Otero de la Roza (National Institute for Nanotechnology, NRC), Edoardo Apra (PNNL).
Future projects: Dynamics on excited-state surfaces, surface hopping, GW/BSE for molecular systems, Spin-flip TDDFT, Non-collinear DFT, spin-orbit TDDFT.
Plane-Wave Density Functional Theory (DFT), Ab Initio Molecular Dynamics, and NWPhys
- Parallel in Time Algorithms. Developers: Eric J. Bylaska (PNNL), Jonathan Q. Weare (University of Chicago), John H. Weare (UCSD).
- New free energy methods based on diffusion Monte-Carlo algorithm. Developers: Eric J. Bylaska (PNNL), Ying Chen (UCSD), John H. Weare (UCSD).
- Dynamic Mean Field Theory (DMFT). Developers: Duo Song (UCSD), Eric J. Bylaska (PNNL), John H. Weare (UCSD).
- Development of new methods to calculate XPS and XANES spectra. Developers: Eric J. Bylaska (PNNL), Niri Govind (PNNL), John Rehr (University of Washington).
- Implementation of electric field gradients and NMR in NWPW Developers: Eric J. Bylaska (PNNL).
- Implementation of the fast multipole method (FMM) in the combined Ab initio molecular dynamics and molecular dynamics (AIMD/MM) code. Developers: Eric J. Bylaska (PNNL).
- Constant pressure ab initio molecular dynamics. Developers: Eric J. Bylaska (PNNL).
- New implementation of the projector augmented wave method in NWPW. Developers: Eric J. Bylaska (PNNL).
- Initial implementation of orbital free DFT in NWPW. Developers: Eric J. Bylaska (PNNL).
- implementation of Hybrid openmp-mpi and offloading intel MIC algorithms in NWPW. Developers: Eric J. Bylaska (PNNL).
Future projects: New NWPhys module development (w/ John Rehr University of Washington) which will include new methods to calculate XPS and XANES spectra. Interface to QWalk Quantum Monte-Carlo Program (w/ Lubos Mitas University of North Carolina).
High-level Coupled-Cluster methods
- Development of multi-reference coupled-cluster capabilities for quasidegenerate systems. Developers: Jiri Pittner (J Heyrovsky Institute of Physical Chemistry), Karol Kowalski (PNNL).
- Electron-affinity/ionization-potential Equation-of-motion Coupled-Cluster methods. Developers: Kiran Bhaskaran-Nair (LSU), Mark Jarrell (LSU), Juana Moreno (LSU), William Shelton (LSU), Karol Kowalski (PNNL).
- Green function Coupled Cluster formalism. Developers: Kiran Bhaskaran-Nair (LSU), Mark Jarrell (LSU), Juana Moreno (LSU), William Shelton (LSU), Karol Kowalski (PNNL).
- Development of Intel MIC implementation of the CCSD(T) approach Developers: Edoardo Apra (PNNL), Michael Klemm (Intel), Karol Kowalski (PNNL).
- Reduced scaling CC formulations based on the Cholesky Decomposition. Developers: Huub van Dam (PNNL), Edoardo Apra (PNNL), Karol Kowalski (PNNL).
Future projects: CC/EOMCC analytical gradients, Intel MIC implementations for iterative CC methods, Multi-reference CC formulations employing incomplete model spaces.
Long-term NWChem development plans:
- Development of new algorithms for hybrid computer architectures including GPU and Intel Xeon Phi computer architectures (NWChem offers already GPU implementations of many-body methods, in 6.5 release we will extend these capabilities to Intel Xeon Phi technology) ,
- Implementation of reduced-scaling methods for electronic structure calculations (local formulations, tensor hypercontractions, resolution-of-identity based approaches),
- Development of novel methodologies for extending temporal scales in ab-initio molecular dynamic and molecular dynamics simulations,
- Approximate electronic structure methods for very large-scale simulations (various semi-empirical methods, order N − N2 DFT algorithms - orbital free DFT),
- Integration and extension of existing capabilities towards predictive models for mesoscale systems (for example, aerosol particles, soil chemistry, biosystems, hormone-cofactor functionality in proteins, ionic liquids in cells, large-scale reactions containing multiple steps).
Please cite the following reference when publishing results obtained with NWChem:
M. Valiev, E.J. Bylaska, N. Govind, K. Kowalski, T.P. Straatsma, H.J.J. van Dam, D. Wang, J. Nieplocha, E. Apra, T.L. Windus, W.A. de Jong, "NWChem: a comprehensive and scalable open-source solution for large scale molecular simulations" Comput. Phys. Commun. 181, 1477 (2010)