http://www.nwchem-sw.org/index.php?title=Special:Contributions&feed=atom&limit=20&target=WikiSysop&year=&month=NWChem - User contributions [en]2019-09-16T03:17:43ZFrom NWChemMediaWiki 1.16.4http://www.nwchem-sw.org/index.php/ChargeCharge2011-03-29T20:36:28Z<p>WikiSysop: </p>
<hr />
<div>=CHARGE=<br />
<br />
This is an optional top-level directive that allows the user to specify the total charge of the system. The form of the directive is as follows:<br />
<br />
CHARGE <real charge default 0><br />
<br />
The default chargeThe charge directive, in conjunction with the charges of atomic nuclei (which can be changed via the geometry input, cf. Section -sec:cart-), determines the total number of electrons in the chemical system. Therefore, a charge n specification removes "n" electrons from the chemical system. Similarly, charge -n adds "n" electrons. is zero if this directive is omitted. An example of a case where the directive would be needed is for a calculation on a doubly charged cation. In such a case, the directive is simply, <br />
<br />
charge 2<br />
<br />
If centers with [[Geometry|fractional charge]] have been specified the net charge of the system should be adjusted to<br />
ensure that there are an integral number of electrons.<br />
<br />
The charge may be changed between tasks, and is used by all wavefunction types. For instance, in order to compute the first two<br />
vertical ionization energies of <math>LiH</math>, one might optimize the geometry of <math>LiH</math> using a UHF SCF wavefunction, and then perform energy calculations at the optimized geometry on <math>LiH^+</math> and <math>LiH^{2+}</math> in turn. This is accomplished with the following input:<br />
<br />
geometry; Li 0 0 0; H 0 0 1.64; end basis; Li library 3-21g; H library 3-21g; end<br />
scf; uhf; singlet; end task scf optimize<br />
charge 1 scf; uhf; doublet; end task scf<br />
charge 2 scf; uhf; singlet; end task scf<br />
<br />
The GEOMETRY, BASIS, and SCF directives are described below ([[Geometry]], [[Basis]] and [[SCF]] respectively) but their intent should be clear. The TASK directive is described above ([[Top-level#TASK|TASK]]).</div>WikiSysophttp://www.nwchem-sw.org/index.php/ChargeCharge2011-03-29T20:33:35Z<p>WikiSysop: </p>
<hr />
<div>=CHARGE=<br />
<br />
This is an optional top-level directive that allows the user to specify the total charge of the system. The form of the directive is as follows:<br />
<br />
CHARGE <real charge default 0><br />
<br />
The default chargeThe charge directive, in conjunction with the charges of atomic nuclei (which can be changed via the geometry input, cf. Section -sec:cart-), determines the total number of electrons in the chemical system. Therefore, a charge n specification removes "n" electrons from the chemical system. Similarly, charge -n adds "n" electrons. is zero if this directive is omitted. An example of a case where the directive would be needed is for a calculation on a doubly charged cation. In such a case, the directive is simply, <br />
<br />
charge 2<br />
<br />
If centers with [[Geometry|fractional charge]] have been specified the net charge of the system should be adjusted to<br />
ensure that there are an integral number of electrons.<br />
<br />
The charge may be changed between tasks, and is used by all wavefunction types. For instance, in order to compute the first two<br />
vertical ionization energies of <math>LiH</math>, one might optimize the geometry of <math>LiH</math> using a UHF SCF wavefunction, and then perform energy calculations at the optimized geometry on <math>LiH^+</math> and <math>LiH^{2+}</math> in turn. This is accomplished with the following input:<br />
<br />
geometry; Li 0 0 0; H 0 0 1.64; end basis; Li library 3-21g; H library 3-21g; end<br />
scf; uhf; singlet; end task scf optimize<br />
charge 1 scf; uhf; doublet; end task scf<br />
charge 2 scf; uhf; singlet; end task scf<br />
<br />
The GEOMETRY, BASIS, and SCF directives are described below ([[Geometry]], [[Basis]] and [[SCF]] respectively) but their intent should be clear. The TASK directive is described above ([[Top-level#TASK|TASK]]).<br />
<br />
PPG</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T22:09:16Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
The format of NWChem parameter is illustrated on the figure below. Fortran code that performs conversion from <br />
AMBER program parameter file format to NWChem is also [[parameter file program|available]].<br />
<br />
[[File:Nwchem-parameter-file.png|frame|Format of NWChem parameter file]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T22:05:39Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
<br />
Figure below illustrates format of NWChem parameter <br />
[[File:Nwchem-parameter-file.png|frame|Format of NWChem parameter file]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T22:02:55Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
Figure below illustrates format of NWChem parameter <br />
[[File:Nwchem-parameter-file.png|caption]]<br />
<br />
Format of NWChem parameter file</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T22:02:40Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
Figure below illustrates format of NWChem parameter <br />
[[File:Nwchem-parameter-file.png|caption]]<br />
Format of NWChem parameter file</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T22:00:59Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
Figure below illustrates format of NWChem parameter <br />
[[File:Nwchem-parameter-file.png|Format of NWChem parameter file]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:59:22Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
Figure below illustrates format of NWChem parameter <br />
[[File:Nwchem-parameter-file.png]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:58:25Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
[[File:Nwchem-parameter-file.png]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:58:10Z<p>WikiSysop: /* Conversion from AMBER program parameter files to NWChem format */</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
[[File:Nwchem-parameter-file.png|right|250px|caption]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:58:00Z<p>WikiSysop: </p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
[[File:Nwchem-parameter-file.pn|right|250px|caption]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:57:14Z<p>WikiSysop: </p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==<br />
<br />
[[File:nwchem-parameter-file.PNG|right|250px|caption]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/File:Nwchem-parameter-file.pngFile:Nwchem-parameter-file.png2010-10-07T21:56:03Z<p>WikiSysop: </p>
<hr />
<div></div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMMQMMM2010-10-07T21:51:29Z<p>WikiSysop: </p>
<hr />
<div>*[[qmmm_introduction|Introduction]]<br />
*[[QMMM_Restart_and_Topology_Files|Topology and Restart Files]]<br />
**[[QMMM_Preparation_Prerequisites|Prerequisites]]<br />
**[[Qmmm_preparation_basic|QM region definition]]<br />
**[[Qmmm_preparation_solvation|Solvation]]<br />
**[[Qmmm_preparation_constraints|Permanent Constraints]]<br />
*[[QMMM_Input_File|Input File]] <br />
**[[QM_Parameters|QM Parameters]] <br />
**[[MM_Parameters|MM Parameters]] <br />
**[[QMMM_Parameters|QM/MM Parameters]] <br />
*[[#QM/MM Single Point Calculations|Single Point Calculations]]<br />
**[[qmmm_sp_energy|Ground State Energy and Gradient]]<br />
**[[QMMM_Excited_States|Excited State Energy]]<br />
**[[qmmm_sp_property|Properties]]<br />
**[[QMMM_ESP|ESP Charge Analysis]]<br />
*[[#Potential Energy Surface Analysis|Potential Energy Surface Analysis]]<br />
**[[qmmm_optimization|Optimization]]<br />
**[[QMMM_Transition_States| Transition States]]<br />
**[[qmmm_freq|Hessians and Frequency]] <br />
**[[qmmm_NEB_Calculations|Reaction Pathway Calculations with NEB]]<br />
*[[QMMM_Dynamics|Dynamics]]<br />
*[[QMMM_Free_Energy|Free Energy Calculations]]<br />
*[[QMMM Appendix|Appendix]]<br />
**[[QMMM Appendix|Conversion from AMBER program parameter files to NWChem]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMM_AppendixQMMM Appendix2010-10-07T21:51:03Z<p>WikiSysop: Created page with '==Conversion from AMBER program parameter files to NWChem format=='</p>
<hr />
<div>==Conversion from AMBER program parameter files to NWChem format==</div>WikiSysophttp://www.nwchem-sw.org/index.php/QMMMQMMM2010-10-07T21:50:54Z<p>WikiSysop: </p>
<hr />
<div>*[[qmmm_introduction|Introduction]]<br />
*[[QMMM_Restart_and_Topology_Files|Topology and Restart Files]]<br />
**[[QMMM_Preparation_Prerequisites|Prerequisites]]<br />
**[[Qmmm_preparation_basic|QM region definition]]<br />
**[[Qmmm_preparation_solvation|Solvation]]<br />
**[[Qmmm_preparation_constraints|Permanent Constraints]]<br />
*[[QMMM_Input_File|Input File]] <br />
**[[QM_Parameters|QM Parameters]] <br />
**[[MM_Parameters|MM Parameters]] <br />
**[[QMMM_Parameters|QM/MM Parameters]] <br />
*[[#QM/MM Single Point Calculations|Single Point Calculations]]<br />
**[[qmmm_sp_energy|Ground State Energy and Gradient]]<br />
**[[QMMM_Excited_States|Excited State Energy]]<br />
**[[qmmm_sp_property|Properties]]<br />
**[[QMMM_ESP|ESP Charge Analysis]]<br />
*[[#Potential Energy Surface Analysis|Potential Energy Surface Analysis]]<br />
**[[qmmm_optimization|Optimization]]<br />
**[[QMMM_Transition_States| Transition States]]<br />
**[[qmmm_freq|Hessians and Frequency]] <br />
**[[qmmm_NEB_Calculations|Reaction Pathway Calculations with NEB]]<br />
*[[QMMM_Dynamics|Dynamics]]<br />
*[[QMMM_Free_Energy|Free Energy Calculations]]<br />
*[[QMMM Appendix|Appendix]]<br />
**[[Appendix|Conversion from AMBER program parameter files to NWChem|]]</div>WikiSysophttp://www.nwchem-sw.org/index.php/ScienceScience2010-09-20T22:02:37Z<p>WikiSysop: /* ECCE Graphical User Interface Provides End-to-End MD Support */</p>
<hr />
<div>__NOTITLE__<br />
=Science with NWChem=<br />
<br />
NWChem used by thousands of researchers worldwide to investigate questions about chemical processes by applying theoretical techniques to predict the structure, properties, and reactivity of chemical and biological species ranging in size from tens to millions of atoms. With NWChem, researchers can tackle molecular systems including biomolecules, nanostructures, [[media:nwchem_actinides.pdf | actinide complexes]], and materials. NWChem offers an extensive array of highly scalable, parallel computational chemistry methods needed to address scientific questions that are relevant to reactive chemical processes occurring in our everyday environment—photosynthesis, protein functions, and combustion, to name a few. They include a multitude of highly correlated methods, density functional theory (DFT) with an extensive set of exchange-correlation functionals, time-dependent density functional theory (TDDFT), plane-wave DFT with exact exchange and Car-Parrinello, molecular dynamics with AMBER and CHARMM force fields, and combinations of them.<br />
<br />
A list of research publications that utilized NWChem can be found [http://www.emsl.pnl.gov/capabilities/computing/nwchem/pubs.jsp here].<br />
<br />
=Ground State DFT Calculations=<br />
<br />
The local basis implementation of DFT in NWChem uses Gaussian atom centered Gaussian type orbitals (GTO) and can be used to study molecular, finite clusters and nanosystems. An exhaustive list of exchange-correlation functionals are supported including: traditional DFT, hybrid functionals, meta-type functionals, range-separated forms, double-hybrid functionals and dispersion corrections. All the available exchange-correlation functionals have associated analytic first derivatives and most functionals have associated second derivatives. Relativistic effects can also be included in DFT/HF calculations via the all-electron spin-free and spin–orbit one-electron Douglas–Kroll–Hess (DKH) method and zeroth-order relativistic approximation (ZORA) as well as through effective core (ECP) and spin–orbit (SO) potentials.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:rochefort_2010.png|<small>''[http://www.rsc.org/delivery/_ArticleLinking/ArticleLinking.cfm?JournalCode=CC&Year=2010&ManuscriptID=b926286e&Iss=17 Adsorption of aminotriazines on graphene using dispersion corrected DFT]</small><br />
File:amity_zeolite.png|<small>''Formyl cation bound to a Bronsted acid site in a zeolite cavity</small><br />
File:tio2_160.png|<small>''[http://www.mrs.org/s_mrs/sec_subscribe.asp?CID=26171&DID=325813&action=detail Ground and excited state properties of TiO<sub>2</sub> rutile]</small><br />
File:waters.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v131/i21/p214103_s1 Dipole polarizabilities of water clusters]</small><br />
File:sic_nano.png|<small>''Ground and excited state properties of SiC nanoclusters</small><br />
</gallery></center><br />
<br />
=Excited State Calculations with TDDFT=<br />
<br />
TDDFT has become an attractive and efficient tool for reliable excited-state calculations involving single excitations in a wide range of systems spanning molecules to materials. However, a major drawback lies in the quality of the exchange-correlation functionals. It is well known that various commonly used functionals are reasonably accurate for valence excited states but fall short in the prediction of charge-transfer (CT) and Rydberg states. The recently developed range-separated functionals have been shown to be very encouraging for CT transitions and outperform their non-range-separated counterparts. Results with these functionals are also comparable with very high levels of theory like EOMCC. In addition to an exhaustive list of commonly used exchange-correlation functionals, several range-separated forms have been recently implemented and extensively validated in NWChem.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:porphyrin_water.png|<small>''[http://pubs.acs.org/doi/full/10.1021/jp902118k Charge transfer excitations in Zinc Porphyrin in aqueous solution]</small><br />
File:ATbasepair.png|<small>''[http://pubs.acs.org/doi/full/10.1021/jp905893v Correct lowest excitation in the AT base pair]</small><br />
File:p2ta.png|<small>''[http://jcp.aip.org/jcpsa6/v132/i15/p154103_s1 Excitation energies for the oligoporphyrin dimer calculated with range-separated TDDFT are in very good agreement with EOMCC and experimental data]</small><br />
File:chromophores.png|<small>''[http://pubs.acs.org/doi/abs/10.1021/ct900231r Optical spectra of TCF-Chromophores]</small><br />
File:ag20.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v132/i19/p194302_s1 Optical properties of silver clusters]</small><br />
</gallery></center><br />
<br />
=High Accuracy Is Scalable to Large Problem Sizes=<br />
<br />
A comprehensive understanding of excited-state processes is crucial in many areas of chemistry, and NWChem has the methodologies capable of providing reliable characterization of excitation energies and excited-state potential energies. Several recent implementations based on the Equation-of-Motion Coupled Cluster (EOMCC) formalism offer a unique chance to address these problems. The largest up-to-date EOMCC calculations demonstrated that taking advantage of ever growing computer power the calculations with the EOMCC methodologies are possible for systems composed of 200-300 correlated electrons.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:P2ta_ver1.png|<small>''[http://jcp.aip.org/jcpsa6/v132/i15/p154103_s1 The EOMCC excitation energies for the oligoporphyrin dimer are in a very good agreement with experimentally inferred values]</small><br />
File:All_trans_beta_carotene_final_dft_ccpvtz.png|<small>''New EOMCC formalisms can be used to study systems of biological relevance (&beta;-carotene)</small><br />
File:Spiro.PNG|<small>''[http://pubs.acs.org/doi/abs/10.1021/jp101761d Highly correlated EOMCC methods accounting for higher-correlation effects can be used to describe electron transfer processes in challenging open-shell systems]</small><br />
File:Tio2_doped.png|<small>''[http://dx.doi.org/10.1016/j.cplett.2009.01.073 New EOMCC models for multiscale approaches provide efficient tool for modeling excite states in true environment]</small><br />
File:Tce_picture_new.png|<small>''Tensor Contraction Engine (TCE) generated EOMCC codes can take advantage of massively parallel compute. The CR-EOMCCSD(T) implementation was shown to scale across 34,008 cores''</small><br />
</gallery></center><br />
<br />
NWChem offers several variants of the EOMCC formalism – from rudimentary EOMCC model with singles and doubles (EOMCCSD) to more sophisticated methods accounting for the effect of triple excitations such as the non-iterative CR-EOMCCSD(T) method and the iterative EOMCCSDT approach and its active-space variant. These accurate formalisms can be used not only for calculations of vertical excitation energies but also to characterize excited-state potential energy surfaces. Various EOMCC approaches can be fully integrated with the multiscale approaches enabling the simulations of the excited-state processes in realistic settings. The correlated methodologies are utilized in this case as high-accuracy drivers for quantum part of the QM/MM and Embedded Clusters formalisms. This functionality provides a unique opportunity of excited-state studies of molecular systems in solution and localized excited states for materials. Another advantage of using TCE generated CC codes is a possibility of utilizing various types of reference functions for closed- and open-shell systems including RHF, ROHF, UHF, and DFT references. Recently developed solvers for the EOMCC theory enable one to deal with excited states of highly diversified configurational structure.<br />
<br />
=Extensive QM/MM Capabilities Are Seamlessly Integrated=<br />
<br />
The new QM/MM module provides a seamless integration between molecular mechanics and most quantum-mechanical theories in NWChem. It boasts an extensive array of capabilities for comprehensive description of large molecular systems for chemistry and biology, including [[Qmmm_sp_energy|ground]] and [[QMMM_Excited_States|excited state]] calculations, [[Qmmm_sp_property|properties]], efficient [[Qmmm_optimization|large scale optimizations]], [[Qmmm_NEB_Calculations|reaction pathways]], [[QMMM_Dynamics|dynamics]], and [[QMMM_Free_Energy|free energy]].<br />
<br />
<center><gallery widths=170px perrow=3 right middle ><br />
File:Large.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v127/i5/p051102_s1 Free energy calculations for chloroform and OH- reaction in aqueous solutions ]''</small><br />
File:Active-site.png |<small>''[http://pubs.acs.org/doi/full/10.1021/jp074853q Studying catalytic of mechanism of protein kinase]''</small><br />
File:QMMM-excited-states.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v125/i21/p211101_s1 Studying excited states in DNA]''</small><br />
</gallery></center><br />
<br />
=High Accuracy Is Achievable for Molecular Properties and Excited States=<br />
<br />
[[File:C60.png|right|200px|Excellent agreement between dipole polarizabilities of C60 calculated with NWChem using LR-CCSD and experimental values]]<br />
<br />
With the recently developed linear-response coupled-cluster module, high-precision calculations for static and dynamic molecular properties are possible. Several levels of theory have been implemented including models based on singles and doubles (LR-CCSD) as well as more sophisticated theories based on the singles, doubles, and triples (LR-CCSDT). The recent studies of static/dynamic polarizabilities for [http://dx.doi.org/10.1063/1.2772853 polyacenes] and [http://dx.doi.org/10.1063/1.3028541 C60 molecule] clearly demonstrate a great potential of the LR-CC theory in retaining the essential part of the correlation effects. The LR-CCSD approach enables also highly-accurate estimates of the [http://dx.doi.org/10.1063/1.3134744 static hyperpolarizabilities for systems where low-rank methods fail].<br />
<br />
=Scalable Plane Wave Has Exact Exchange=<br />
<br />
[[File:Bylaska_figure1b.JPG|left|150px|Unique relativistic Car-Parrinello simulations uranyl in water show excellent agreement with available experimental data.]]<br />
[[File:Bylaska-Figure3.PNG|right|200px|caption]]<br />
The NWChem highly scalable Plane-Wave module now enables users to use exact exchange and the Self-Interaction Correction (SIC) within its framework for complex molecular, liquid, and solid-state systems. In addition, great improvements in parallel performance allow users to study larger systems at longer time scales with a faster time-to-solution. Scaling to 24,000 processors has been demonstrated for a 160 atom hematite cluster using density functional theory with exact exchange.<br />
<br />
NWChem’s Car-Parrinello simulation capability, incorporating relativistic effects, was used to [http://jcp.aip.org/resource/1/jcpsa6/v128/i12/p124507_s1 model the behavior of uranium in complex solution environments] and on iron-oxide surfaces as well as clay layers.<br />
<br />
=Molecular Dynamics of Complex Biomolecular Systems=<br />
<br />
[[File:OprF.png|right|200px|Outer membrane OprF of Pseudomonas aeruginosa in a lipopolysaccharide membrane.]]<br />
<br />
The Molecular Dynamics (MD) module of NWChem is designed for the simulation of biomolecular systems, with special features that facilitate the setup of complex systems such as memprane proteins embedded in complex microbial membranes consisting. These capabilities have been used to perform the first ever computer simulations of outer membrane proteins in lipopolysaccharide membranes, in a study of the [http://dx.doi.org/10.1002/prot.22165 OprF outer membrane protein of Pseudomonas aeruginosa.]<br />
<br />
=ECCE Graphical User Interface Provides End-to-End MD Support=<br />
<br />
[[File:Ecce.jpg|left|200px|Ecce’s end-to-end support for molecular dynamics calculations in NWChem greatly increases the ease of use.]]<br />
<br />
[http://ecce.pnl.gov/ ECCE] (Extensible Computational Chemistry Environment) provides full end-to-end support for the molecular dynamics module in NWChem. Through the graphical interface, users can set up their calculations, submit them to their computer platforms, and visualize the results on-the-fly. <br />
<br />
ECCE provides an easy to use graphical user interface that enables novice users to set up quantum chemistry calculations, submit them on large supercomputers, and analyze and visualize the calculated data. The next release will have the capability to build solid state materials and surfaces.</div>WikiSysophttp://www.nwchem-sw.org/index.php/ScienceScience2010-09-20T22:02:27Z<p>WikiSysop: /* Ecce Graphical User Interface Provides End-to-End MD Support */</p>
<hr />
<div>__NOTITLE__<br />
=Science with NWChem=<br />
<br />
NWChem used by thousands of researchers worldwide to investigate questions about chemical processes by applying theoretical techniques to predict the structure, properties, and reactivity of chemical and biological species ranging in size from tens to millions of atoms. With NWChem, researchers can tackle molecular systems including biomolecules, nanostructures, [[media:nwchem_actinides.pdf | actinide complexes]], and materials. NWChem offers an extensive array of highly scalable, parallel computational chemistry methods needed to address scientific questions that are relevant to reactive chemical processes occurring in our everyday environment—photosynthesis, protein functions, and combustion, to name a few. They include a multitude of highly correlated methods, density functional theory (DFT) with an extensive set of exchange-correlation functionals, time-dependent density functional theory (TDDFT), plane-wave DFT with exact exchange and Car-Parrinello, molecular dynamics with AMBER and CHARMM force fields, and combinations of them.<br />
<br />
A list of research publications that utilized NWChem can be found [http://www.emsl.pnl.gov/capabilities/computing/nwchem/pubs.jsp here].<br />
<br />
=Ground State DFT Calculations=<br />
<br />
The local basis implementation of DFT in NWChem uses Gaussian atom centered Gaussian type orbitals (GTO) and can be used to study molecular, finite clusters and nanosystems. An exhaustive list of exchange-correlation functionals are supported including: traditional DFT, hybrid functionals, meta-type functionals, range-separated forms, double-hybrid functionals and dispersion corrections. All the available exchange-correlation functionals have associated analytic first derivatives and most functionals have associated second derivatives. Relativistic effects can also be included in DFT/HF calculations via the all-electron spin-free and spin–orbit one-electron Douglas–Kroll–Hess (DKH) method and zeroth-order relativistic approximation (ZORA) as well as through effective core (ECP) and spin–orbit (SO) potentials.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:rochefort_2010.png|<small>''[http://www.rsc.org/delivery/_ArticleLinking/ArticleLinking.cfm?JournalCode=CC&Year=2010&ManuscriptID=b926286e&Iss=17 Adsorption of aminotriazines on graphene using dispersion corrected DFT]</small><br />
File:amity_zeolite.png|<small>''Formyl cation bound to a Bronsted acid site in a zeolite cavity</small><br />
File:tio2_160.png|<small>''[http://www.mrs.org/s_mrs/sec_subscribe.asp?CID=26171&DID=325813&action=detail Ground and excited state properties of TiO<sub>2</sub> rutile]</small><br />
File:waters.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v131/i21/p214103_s1 Dipole polarizabilities of water clusters]</small><br />
File:sic_nano.png|<small>''Ground and excited state properties of SiC nanoclusters</small><br />
</gallery></center><br />
<br />
=Excited State Calculations with TDDFT=<br />
<br />
TDDFT has become an attractive and efficient tool for reliable excited-state calculations involving single excitations in a wide range of systems spanning molecules to materials. However, a major drawback lies in the quality of the exchange-correlation functionals. It is well known that various commonly used functionals are reasonably accurate for valence excited states but fall short in the prediction of charge-transfer (CT) and Rydberg states. The recently developed range-separated functionals have been shown to be very encouraging for CT transitions and outperform their non-range-separated counterparts. Results with these functionals are also comparable with very high levels of theory like EOMCC. In addition to an exhaustive list of commonly used exchange-correlation functionals, several range-separated forms have been recently implemented and extensively validated in NWChem.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:porphyrin_water.png|<small>''[http://pubs.acs.org/doi/full/10.1021/jp902118k Charge transfer excitations in Zinc Porphyrin in aqueous solution]</small><br />
File:ATbasepair.png|<small>''[http://pubs.acs.org/doi/full/10.1021/jp905893v Correct lowest excitation in the AT base pair]</small><br />
File:p2ta.png|<small>''[http://jcp.aip.org/jcpsa6/v132/i15/p154103_s1 Excitation energies for the oligoporphyrin dimer calculated with range-separated TDDFT are in very good agreement with EOMCC and experimental data]</small><br />
File:chromophores.png|<small>''[http://pubs.acs.org/doi/abs/10.1021/ct900231r Optical spectra of TCF-Chromophores]</small><br />
File:ag20.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v132/i19/p194302_s1 Optical properties of silver clusters]</small><br />
</gallery></center><br />
<br />
=High Accuracy Is Scalable to Large Problem Sizes=<br />
<br />
A comprehensive understanding of excited-state processes is crucial in many areas of chemistry, and NWChem has the methodologies capable of providing reliable characterization of excitation energies and excited-state potential energies. Several recent implementations based on the Equation-of-Motion Coupled Cluster (EOMCC) formalism offer a unique chance to address these problems. The largest up-to-date EOMCC calculations demonstrated that taking advantage of ever growing computer power the calculations with the EOMCC methodologies are possible for systems composed of 200-300 correlated electrons.<br />
<br />
<center><gallery widths=170px perrow=5><br />
File:P2ta_ver1.png|<small>''[http://jcp.aip.org/jcpsa6/v132/i15/p154103_s1 The EOMCC excitation energies for the oligoporphyrin dimer are in a very good agreement with experimentally inferred values]</small><br />
File:All_trans_beta_carotene_final_dft_ccpvtz.png|<small>''New EOMCC formalisms can be used to study systems of biological relevance (&beta;-carotene)</small><br />
File:Spiro.PNG|<small>''[http://pubs.acs.org/doi/abs/10.1021/jp101761d Highly correlated EOMCC methods accounting for higher-correlation effects can be used to describe electron transfer processes in challenging open-shell systems]</small><br />
File:Tio2_doped.png|<small>''[http://dx.doi.org/10.1016/j.cplett.2009.01.073 New EOMCC models for multiscale approaches provide efficient tool for modeling excite states in true environment]</small><br />
File:Tce_picture_new.png|<small>''Tensor Contraction Engine (TCE) generated EOMCC codes can take advantage of massively parallel compute. The CR-EOMCCSD(T) implementation was shown to scale across 34,008 cores''</small><br />
</gallery></center><br />
<br />
NWChem offers several variants of the EOMCC formalism – from rudimentary EOMCC model with singles and doubles (EOMCCSD) to more sophisticated methods accounting for the effect of triple excitations such as the non-iterative CR-EOMCCSD(T) method and the iterative EOMCCSDT approach and its active-space variant. These accurate formalisms can be used not only for calculations of vertical excitation energies but also to characterize excited-state potential energy surfaces. Various EOMCC approaches can be fully integrated with the multiscale approaches enabling the simulations of the excited-state processes in realistic settings. The correlated methodologies are utilized in this case as high-accuracy drivers for quantum part of the QM/MM and Embedded Clusters formalisms. This functionality provides a unique opportunity of excited-state studies of molecular systems in solution and localized excited states for materials. Another advantage of using TCE generated CC codes is a possibility of utilizing various types of reference functions for closed- and open-shell systems including RHF, ROHF, UHF, and DFT references. Recently developed solvers for the EOMCC theory enable one to deal with excited states of highly diversified configurational structure.<br />
<br />
=Extensive QM/MM Capabilities Are Seamlessly Integrated=<br />
<br />
The new QM/MM module provides a seamless integration between molecular mechanics and most quantum-mechanical theories in NWChem. It boasts an extensive array of capabilities for comprehensive description of large molecular systems for chemistry and biology, including [[Qmmm_sp_energy|ground]] and [[QMMM_Excited_States|excited state]] calculations, [[Qmmm_sp_property|properties]], efficient [[Qmmm_optimization|large scale optimizations]], [[Qmmm_NEB_Calculations|reaction pathways]], [[QMMM_Dynamics|dynamics]], and [[QMMM_Free_Energy|free energy]].<br />
<br />
<center><gallery widths=170px perrow=3 right middle ><br />
File:Large.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v127/i5/p051102_s1 Free energy calculations for chloroform and OH- reaction in aqueous solutions ]''</small><br />
File:Active-site.png |<small>''[http://pubs.acs.org/doi/full/10.1021/jp074853q Studying catalytic of mechanism of protein kinase]''</small><br />
File:QMMM-excited-states.png|<small>''[http://jcp.aip.org/resource/1/jcpsa6/v125/i21/p211101_s1 Studying excited states in DNA]''</small><br />
</gallery></center><br />
<br />
=High Accuracy Is Achievable for Molecular Properties and Excited States=<br />
<br />
[[File:C60.png|right|200px|Excellent agreement between dipole polarizabilities of C60 calculated with NWChem using LR-CCSD and experimental values]]<br />
<br />
With the recently developed linear-response coupled-cluster module, high-precision calculations for static and dynamic molecular properties are possible. Several levels of theory have been implemented including models based on singles and doubles (LR-CCSD) as well as more sophisticated theories based on the singles, doubles, and triples (LR-CCSDT). The recent studies of static/dynamic polarizabilities for [http://dx.doi.org/10.1063/1.2772853 polyacenes] and [http://dx.doi.org/10.1063/1.3028541 C60 molecule] clearly demonstrate a great potential of the LR-CC theory in retaining the essential part of the correlation effects. The LR-CCSD approach enables also highly-accurate estimates of the [http://dx.doi.org/10.1063/1.3134744 static hyperpolarizabilities for systems where low-rank methods fail].<br />
<br />
=Scalable Plane Wave Has Exact Exchange=<br />
<br />
[[File:Bylaska_figure1b.JPG|left|150px|Unique relativistic Car-Parrinello simulations uranyl in water show excellent agreement with available experimental data.]]<br />
[[File:Bylaska-Figure3.PNG|right|200px|caption]]<br />
The NWChem highly scalable Plane-Wave module now enables users to use exact exchange and the Self-Interaction Correction (SIC) within its framework for complex molecular, liquid, and solid-state systems. In addition, great improvements in parallel performance allow users to study larger systems at longer time scales with a faster time-to-solution. Scaling to 24,000 processors has been demonstrated for a 160 atom hematite cluster using density functional theory with exact exchange.<br />
<br />
NWChem’s Car-Parrinello simulation capability, incorporating relativistic effects, was used to [http://jcp.aip.org/resource/1/jcpsa6/v128/i12/p124507_s1 model the behavior of uranium in complex solution environments] and on iron-oxide surfaces as well as clay layers.<br />
<br />
=Molecular Dynamics of Complex Biomolecular Systems=<br />
<br />
[[File:OprF.png|right|200px|Outer membrane OprF of Pseudomonas aeruginosa in a lipopolysaccharide membrane.]]<br />
<br />
The Molecular Dynamics (MD) module of NWChem is designed for the simulation of biomolecular systems, with special features that facilitate the setup of complex systems such as memprane proteins embedded in complex microbial membranes consisting. These capabilities have been used to perform the first ever computer simulations of outer membrane proteins in lipopolysaccharide membranes, in a study of the [http://dx.doi.org/10.1002/prot.22165 OprF outer membrane protein of Pseudomonas aeruginosa.]<br />
<br />
=ECCE Graphical User Interface Provides End-to-End MD Support=<br />
<br />
[[File:Ecce.jpg|left|200px|Ecce’s end-to-end support for molecular dynamics calculations in NWChem greatly increases the ease of use.]]<br />
<br />
[http://ecce.pnl.gov/ECCE] (Extensible Computational Chemistry Environment) provides full end-to-end support for the molecular dynamics module in NWChem. Through the graphical interface, users can set up their calculations, submit them to their computer platforms, and visualize the results on-the-fly. <br />
<br />
Ecce provides an easy to use graphical user interface that enables novice users to set up quantum chemistry calculations, submit them on large supercomputers, and analyze and visualize the calculated data. The next release will have the capability to build solid state materials and surfaces.</div>WikiSysophttp://www.nwchem-sw.org/index.php/BenchmarksBenchmarks2010-09-18T20:55:51Z<p>WikiSysop: /* Parallel performance of Ab initio Molecular Dynamics using plane waves */</p>
<hr />
<div>__NOTITLE__<br />
<br />
=Benchmarks performed with NWChem=<br />
<br />
This page contains a suite of benchmarks performed with NWChem. The benchmarks include a variety of computational chemistry methods on a variety of high performance computing platforms. The list of benchmarks available will evolve continuously as new data becomes available. If you have benchmark information you would like to add for your computing system, please contact one of the developers.<br />
<br />
=Hybrid density functional calculation on the C<sub>240</sub> Buckyball=<br />
<br />
Performance of the Gaussian basis set DFT module in NWChem. This calculation involved performing a PBE0 calculation (in direct mode) on the on C<sub>240</sub> system with the 6-31G* basis set (3600 basis functions). These calculations were performed on the Chinook supercomputer located at PNNL. Timings are per step for the various components. The [[Media:input_c240_pbe0.nw|input file]] is available.<br />
<br />
[[File:dft-scaling-c240-pbe02.png|center|300px]]<br />
<br />
=Parallel performance of ''Ab initio'' Molecular Dynamics using plane waves=<br />
<br />
[[file:PWScaling.png|left|220px |thumb|AIMD Parallel timings for <math>UO_2^{2+}</math>+122<math>H_2</math>O. These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWEXScaling.png|right|220px|thumb|Exact exchange timings – 80 atom cell of hematite (cutoff energy=100Ry). These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWMDScaling.png|center|220px| thumb|AIMD and AIMD/MM Parallel Timings for <math>Zn^{2+}</math>+64<math>H_2</math>O (unit cell parameters SC=12.4 Angs. and cutoff energy =100Ry). These calculations were performed on the Chinook HP computer system at MSCF EMSL, PNNL.]]<br />
<br />
=Parallel performance of the CR-EOMCCSD(T) method (triples part)= <br />
<br />
An example of the scalability of the triples part of the [http://dx.doi.org/10.1063/1.3385315 CR-EOMCCSD(T) approach] for Green Fluorescent Protein Chromophore (GFPC)<br />
described by cc-pVTZ basis set (648 basis functions) as obtained from NWChem. Timings were determined from calculations on the Franklin Cray-XT4 computer system at NERSC.<br />
See the [[Media:input_gfpc.nw| input file]] for details.<br />
<br />
[[File:creomccsd_t.png|center|300px| ]]<br />
<br />
=Timings of CCSD/EOMCCSD for the oligoporphyrin dimer =<br />
<br />
CCSD/EOMCCSD timings for oligoporphyrin dimer (942 basis functions, 270 correlated electrons, D2h symmetry, excited-state calculations were <br />
performed for state of b1g symmetry, in all test calculation convergence threshold was relaxed, 1024 cores were used). See the [[Media:input_p2ta.nw| input file]] for details.<br />
<br />
--------------------------------------------------------<br />
Iter Residuum Correlation Cpu Wall<br />
--------------------------------------------------------<br />
1 0.7187071521175 -7.9406033677717 640.9 807.7<br />
......<br />
MICROCYCLE DIIS UPDATE: 10 5<br />
11 0.0009737920958 -7.9953441809574 691.1 822.2<br />
--------------------------------------------------------<br />
Iterations converged<br />
CCSD correlation energy / hartree = -7.995344180957357<br />
CCSD total energy / hartree = -2418.570838364838890<br />
<br />
EOM-CCSD right-hand side iterations<br />
--------------------------------------------------------------<br />
Residuum Omega / hartree Omega / eV Cpu Wall<br />
--------------------------------------------------------------<br />
......<br />
Iteration 2 using 6 trial vectors<br />
0.1584284659595 0.0882389635508 2.40111 865.3 1041.2<br />
Iteration 3 using 7 trial vectors<br />
0.0575982107592 0.0810948687618 2.20670 918.0 1042.2<br />
<br />
=Current developments for high accuracy: GPGPU and alternative task schedulers=<br />
<br />
Currently various development efforts are underway for high accuracy methods that will be available in future releases of NWChem. The examples below shows the first results of the performance of the triples part of Reg-CCSD(T) on GPGPUs (left two examples) and of using alternative task schedules for the iterative CCSD and EOMCCSD.<br />
<br />
<gallery widths=170px perrow=5><br />
File:gpu_scaling_spiro.png|<small>''Scalability of the triples part of the Reg-CCSD(T) approach for Spiro cation described by the Sadlej's TZ basis set (POL1).<br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:gpu_speedup_uracil.png|<small>''Speedup of GPU over CPU of the (T) part of the (T) part of the Reg-CCSD(T) approach as a function of the tile size for the uracil molecule. <br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:ccsd_eomccsd_new.png|<small>''Comparison of the CCSD/EOMCCSD iteration times for BacterioChlorophyll (BChl) for various tile sizes. Calculations were performed for 3-21G basis set (503 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores).</small><br />
File:bchl_6_311G_ccsd.png|<small>''Time per CCSD iteration for BChl in 6-311G basis set (733 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores) as a function of tile size.</small><br />
File:ccsd_scaling_ic.png|<small>''Scalability of the CCSD code for BChl in 6-311G basis set (733 basis functions; tilesize=40, C1 symmetry, 240 correlated electrons).</small><br />
</gallery></div>WikiSysophttp://www.nwchem-sw.org/index.php/BenchmarksBenchmarks2010-09-18T20:54:42Z<p>WikiSysop: /* Parallel performance of Ab initio Molecular Dynamics using plane waves */</p>
<hr />
<div>__NOTITLE__<br />
<br />
=Benchmarks performed with NWChem=<br />
<br />
This page contains a suite of benchmarks performed with NWChem. The benchmarks include a variety of computational chemistry methods on a variety of high performance computing platforms. The list of benchmarks available will evolve continuously as new data becomes available. If you have benchmark information you would like to add for your computing system, please contact one of the developers.<br />
<br />
=Hybrid density functional calculation on the C<sub>240</sub> Buckyball=<br />
<br />
Performance of the Gaussian basis set DFT module in NWChem. This calculation involved performing a PBE0 calculation (in direct mode) on the on C<sub>240</sub> system with the 6-31G* basis set (3600 basis functions). These calculations were performed on the Chinook supercomputer located at PNNL. Timings are per step for the various components. The [[Media:input_c240_pbe0.nw|input file]] is available.<br />
<br />
[[File:dft-scaling-c240-pbe02.png|center|300px]]<br />
<br />
=Parallel performance of ''Ab initio'' Molecular Dynamics using plane waves=<br />
<br />
[[file:PWScaling.png|left|220px |thumb|AIMD Parallel timings for UO<math>_2^{2+}</math>+122H<math>_2</math>O. These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWEXScaling.png|right|220px|thumb|Exact exchange timings – 80 atom cell of hematite (cutoff energy=100Ry). These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWMDScaling.png|center|220px| thumb|AIMD and AIMD/MM Parallel Timings for Zn<math>^{2+}</math>+64H<math>_2</math>O (unit cell parameters SC=12.4 Angs. and cutoff energy =100Ry). These calculations were performed on the Chinook HP computer system at MSCF EMSL, PNNL.]]<br />
<br />
=Parallel performance of the CR-EOMCCSD(T) method (triples part)= <br />
<br />
An example of the scalability of the triples part of the [http://dx.doi.org/10.1063/1.3385315 CR-EOMCCSD(T) approach] for Green Fluorescent Protein Chromophore (GFPC)<br />
described by cc-pVTZ basis set (648 basis functions) as obtained from NWChem. Timings were determined from calculations on the Franklin Cray-XT4 computer system at NERSC.<br />
See the [[Media:input_gfpc.nw| input file]] for details.<br />
<br />
[[File:creomccsd_t.png|center|300px| ]]<br />
<br />
=Timings of CCSD/EOMCCSD for the oligoporphyrin dimer =<br />
<br />
CCSD/EOMCCSD timings for oligoporphyrin dimer (942 basis functions, 270 correlated electrons, D2h symmetry, excited-state calculations were <br />
performed for state of b1g symmetry, in all test calculation convergence threshold was relaxed, 1024 cores were used). See the [[Media:input_p2ta.nw| input file]] for details.<br />
<br />
--------------------------------------------------------<br />
Iter Residuum Correlation Cpu Wall<br />
--------------------------------------------------------<br />
1 0.7187071521175 -7.9406033677717 640.9 807.7<br />
......<br />
MICROCYCLE DIIS UPDATE: 10 5<br />
11 0.0009737920958 -7.9953441809574 691.1 822.2<br />
--------------------------------------------------------<br />
Iterations converged<br />
CCSD correlation energy / hartree = -7.995344180957357<br />
CCSD total energy / hartree = -2418.570838364838890<br />
<br />
EOM-CCSD right-hand side iterations<br />
--------------------------------------------------------------<br />
Residuum Omega / hartree Omega / eV Cpu Wall<br />
--------------------------------------------------------------<br />
......<br />
Iteration 2 using 6 trial vectors<br />
0.1584284659595 0.0882389635508 2.40111 865.3 1041.2<br />
Iteration 3 using 7 trial vectors<br />
0.0575982107592 0.0810948687618 2.20670 918.0 1042.2<br />
<br />
=Current developments for high accuracy: GPGPU and alternative task schedulers=<br />
<br />
Currently various development efforts are underway for high accuracy methods that will be available in future releases of NWChem. The examples below shows the first results of the performance of the triples part of Reg-CCSD(T) on GPGPUs (left two examples) and of using alternative task schedules for the iterative CCSD and EOMCCSD.<br />
<br />
<gallery widths=170px perrow=5><br />
File:gpu_scaling_spiro.png|<small>''Scalability of the triples part of the Reg-CCSD(T) approach for Spiro cation described by the Sadlej's TZ basis set (POL1).<br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:gpu_speedup_uracil.png|<small>''Speedup of GPU over CPU of the (T) part of the (T) part of the Reg-CCSD(T) approach as a function of the tile size for the uracil molecule. <br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:ccsd_eomccsd_new.png|<small>''Comparison of the CCSD/EOMCCSD iteration times for BacterioChlorophyll (BChl) for various tile sizes. Calculations were performed for 3-21G basis set (503 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores).</small><br />
File:bchl_6_311G_ccsd.png|<small>''Time per CCSD iteration for BChl in 6-311G basis set (733 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores) as a function of tile size.</small><br />
File:ccsd_scaling_ic.png|<small>''Scalability of the CCSD code for BChl in 6-311G basis set (733 basis functions; tilesize=40, C1 symmetry, 240 correlated electrons).</small><br />
</gallery></div>WikiSysophttp://www.nwchem-sw.org/index.php/BenchmarksBenchmarks2010-09-18T20:53:48Z<p>WikiSysop: /* Parallel performance of Ab initio Molecular Dynamics using plane waves */</p>
<hr />
<div>__NOTITLE__<br />
<br />
=Benchmarks performed with NWChem=<br />
<br />
This page contains a suite of benchmarks performed with NWChem. The benchmarks include a variety of computational chemistry methods on a variety of high performance computing platforms. The list of benchmarks available will evolve continuously as new data becomes available. If you have benchmark information you would like to add for your computing system, please contact one of the developers.<br />
<br />
=Hybrid density functional calculation on the C<sub>240</sub> Buckyball=<br />
<br />
Performance of the Gaussian basis set DFT module in NWChem. This calculation involved performing a PBE0 calculation (in direct mode) on the on C<sub>240</sub> system with the 6-31G* basis set (3600 basis functions). These calculations were performed on the Chinook supercomputer located at PNNL. Timings are per step for the various components. The [[Media:input_c240_pbe0.nw|input file]] is available.<br />
<br />
[[File:dft-scaling-c240-pbe02.png|center|300px]]<br />
<br />
=Parallel performance of ''Ab initio'' Molecular Dynamics using plane waves=<br />
<br />
[[file:PWScaling.png|left|200px |thumb|AIMD Parallel timings for UO<math>_2^{2+}</math>+122H<math>_2</math>O. These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWEXScaling.png|right|200px|thumb|Exact exchange timings – 80 atom cell of hematite (cutoff energy=100Ry). These calculations were performed on the Franklin Cray-XT4 computer system at NERSC.]]<br />
[[file:PWMDScaling.png|center|200px| thumb|AIMD and AIMD/MM Parallel Timings for Zn<math>^{2+}</math>+64H<math>_2</math>O (unit cell parameters SC=12.4 Angs. and cutoff energy =100Ry). These calculations were performed on the Chinook HP computer system at MSCF EMSL, PNNL.]]<br />
<br />
=Parallel performance of the CR-EOMCCSD(T) method (triples part)= <br />
<br />
An example of the scalability of the triples part of the [http://dx.doi.org/10.1063/1.3385315 CR-EOMCCSD(T) approach] for Green Fluorescent Protein Chromophore (GFPC)<br />
described by cc-pVTZ basis set (648 basis functions) as obtained from NWChem. Timings were determined from calculations on the Franklin Cray-XT4 computer system at NERSC.<br />
See the [[Media:input_gfpc.nw| input file]] for details.<br />
<br />
[[File:creomccsd_t.png|center|300px| ]]<br />
<br />
=Timings of CCSD/EOMCCSD for the oligoporphyrin dimer =<br />
<br />
CCSD/EOMCCSD timings for oligoporphyrin dimer (942 basis functions, 270 correlated electrons, D2h symmetry, excited-state calculations were <br />
performed for state of b1g symmetry, in all test calculation convergence threshold was relaxed, 1024 cores were used). See the [[Media:input_p2ta.nw| input file]] for details.<br />
<br />
--------------------------------------------------------<br />
Iter Residuum Correlation Cpu Wall<br />
--------------------------------------------------------<br />
1 0.7187071521175 -7.9406033677717 640.9 807.7<br />
......<br />
MICROCYCLE DIIS UPDATE: 10 5<br />
11 0.0009737920958 -7.9953441809574 691.1 822.2<br />
--------------------------------------------------------<br />
Iterations converged<br />
CCSD correlation energy / hartree = -7.995344180957357<br />
CCSD total energy / hartree = -2418.570838364838890<br />
<br />
EOM-CCSD right-hand side iterations<br />
--------------------------------------------------------------<br />
Residuum Omega / hartree Omega / eV Cpu Wall<br />
--------------------------------------------------------------<br />
......<br />
Iteration 2 using 6 trial vectors<br />
0.1584284659595 0.0882389635508 2.40111 865.3 1041.2<br />
Iteration 3 using 7 trial vectors<br />
0.0575982107592 0.0810948687618 2.20670 918.0 1042.2<br />
<br />
=Current developments for high accuracy: GPGPU and alternative task schedulers=<br />
<br />
Currently various development efforts are underway for high accuracy methods that will be available in future releases of NWChem. The examples below shows the first results of the performance of the triples part of Reg-CCSD(T) on GPGPUs (left two examples) and of using alternative task schedules for the iterative CCSD and EOMCCSD.<br />
<br />
<gallery widths=170px perrow=5><br />
File:gpu_scaling_spiro.png|<small>''Scalability of the triples part of the Reg-CCSD(T) approach for Spiro cation described by the Sadlej's TZ basis set (POL1).<br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:gpu_speedup_uracil.png|<small>''Speedup of GPU over CPU of the (T) part of the (T) part of the Reg-CCSD(T) approach as a function of the tile size for the uracil molecule. <br />
The calculations were performed using Barracuda cluster at EMSL.</small><br />
File:ccsd_eomccsd_new.png|<small>''Comparison of the CCSD/EOMCCSD iteration times for BacterioChlorophyll (BChl) for various tile sizes. Calculations were performed for 3-21G basis set (503 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores).</small><br />
File:bchl_6_311G_ccsd.png|<small>''Time per CCSD iteration for BChl in 6-311G basis set (733 basis functions, C1 symmetry, 240 correlated electrons, 1020 cores) as a function of tile size.</small><br />
File:ccsd_scaling_ic.png|<small>''Scalability of the CCSD code for BChl in 6-311G basis set (733 basis functions; tilesize=40, C1 symmetry, 240 correlated electrons).</small><br />
</gallery></div>WikiSysop