PSI3 is a program system and development platform for ab initio molecular electronic structure computations.The package includes mature programming interfaces for parsing user input, accessing commonly used data such as basis-set information or molecular orbital coefficients, and retrieving and storing binary data (with no software limitations on file sizes or file-system-sizes), especially multi-index quantities such as electron repulsion integrals. This platform is useful for the rapid implementation of both standard quantum chemical methods, as well as the development of new models. Features that have already been implemented include Hartree-Fock, multiconfigurational self-consistentfield, second-order Møller-Plesset perturbation theory, coupled cluster, and configuration interaction wave functions. Distinctive capabilities include the ability to employ Gaussian basis functions with arbitrary angular momentum levels; linear R12 second-order perturbation theory; coupled cluster frequency-dependent response properties, including dipole polarizabilities and optical rotation; and diagonal Born-Oppenheimer corrections with correlated wave functions. This article describes the programming infrastructure and main features of the package. PSI3 is available free of charge through the open-source, GNU General Public License.

Linear algebra is a powerful and proven tool in web search. Techniques, such as the PageRank algorithm of Brin and Page and the HITS algorithm of Kleinberg, score web pages based on the principal eigenvector (or singular vector) of a particular non-negative matrix that captures the hyperlink structure of the web graph. We propose and test a new methodology that uses multilinear algebra to elicit more information from a higher-order representation of the hyperlink graph. We start by labeling the edges in our graph with the anchor text of the hyperlinks so that the associated linear algebra representation is a sparse, three-way tensor. The first two dimensions of the tensor represent the web pages while the third dimension adds the anchor text. We then use the rank-1 factors of a multilinear PARAFAC tensor decomposition, which are akin to singular vectors of the SVD, to automatically identify topics in the collection along with the associated authoritative web pages.3 4548

Homoleptic binary cobalt carbonyls with multiple cobalt−cobalt bonds have been examined theoretically using established levels of density functional methodology. These species include 19 structures ranging from the experimentally well characterized dibridged (CO)3Co(CO)2Co(CO)3 to the proposed monobridged (CO)2Co(CO)Co(CO)2 structure with a formal quadruple bond. Consistent with experiment, three energetically low-lying equilibrium structures of Co2(CO)8 were found, of C 2 v (dibridged), D 3 d (unbridged), and D 2 d (unbridged) symmetry. For Co2(CO)8, the BP86 method predicts the dibridged structure to lie 3.7 kcal/mol below the D 2 d structure and 6.3 kcal/mol below the D 3 d structure. The D 2 d and D 3 d structures thus have the opposite energetic ordering of that deduced from experiment by Sweany and Brown. A satisfactory harmony between theoretical and experimental vibrational frequencies and IR intensities is found, although the D 2 d and D 3 d structures are essentially indistinguishable in this regard. For Co2(CO)7 the unbridged asymmetric structure suggested by Sweany and Brown is confirmed with the BP86 method, and with perhaps one exception the vibrational features agree well for theory and experiment. For Co2(CO)6 only one vibrational feature has been assigned from experiment, but this band (2011 cm-1) fits very well with BP86 predictions for the monobridged D 2 d symmetry structure with a formal Co⋮Co triple bond. For the Co2(CO)5 molecule, for which no experimental results exist, the most interesting structure is the monobridged closed-shell singlet with a very short (2.17 Å) cobalt−cobalt bond, to which we assign a formal bond order of four. Potential energy distributions have been analyzed to identify the principal vibrations with cobalt−cobalt stretching contributions. The condensed phase Raman analysis by Onaka and Shriver of the Co−Co stretches for the three known isomers of Co2(CO)8 is remarkably consistent with the present predictions for the gas-phase species. Prospects for the synthesis of these and related dicobalt compounds are discussed.

The problematic SiC2 barrier to linearity is investigated in a benchmark study of one-electron basis set convergence properties of both the conventional and linear R12/A formulations of second-order Møller–Plesset (MP2) perturbation theory. A procedure for computational molecular partial-wave expansions is constructed and applied to the T-shaped and linear forms of SiC2. The largest basis set used [Si(22s17p14d6f5g2h2i1k)/C(19s14p14d6f5g2h2i1k)] included functions of orbital angular momentum as large as l=7 (k), and systematic saturation was performed through l=6 (i). With respect to angular momentum l, correlation energy increments are found to decay in accord with analytical models that suggest (l+1/2)−6 and (l+1/2)−4 functional forms for the R12/A and conventional methods, respectively. A benchmark complete basis set (CBS) limit for the second-order correlation contribution to the SiC2 barrier to linearity, 5.66 kcal mol−1, was determined via MP2-R12/A partial-wave expansions. Conventional MP2 calculations, using both the standard cc-pV6Z and the [Si(22s17p14d6f5g2h2i1k)/C(19s14p14d6f5g2h2i1k)] basis sets, underestimate MP2 correlation energies by at least 3 kcal mol−1, while the barrier is underestimated by at least 0.1 kcal mol−1. Both X−3 cc-pVXZ extrapolations and partial-wave extrapolations greatly improve conventional correlation energies, with the cc-pVXZ extrapolated barrier in error by only 0.07 kcal mol−1. While the absolute accuracy of the conventional partial-wave extrapolations is substantially better than the cc-pVXZ extrapolated values, unbalanced errors result in an overestimation of the barrier by nearly 0.2 kcal mol−1. The CBS-limit MP2 contribution is combined via a focal-point analysis with conventional coupled cluster computations through triple excitations (CCSDT), resulting in an inferred CBS CCSDT barrier of 5.45 kcal mol−1 after accounting for core correlation and relativistic effects. The critical question of post-CCSDT corrections is approached through explicit coupled cluster computations perturbatively accounting for connected quadruple excitations [BD(TQ) and CCSD(2)], as well as shifted [2,1] Padé approximants of MPn series and continued fraction and quadratic Padé approximants of coupled-cluster series. The best available post-CCSDT correction, extracted from BD(TQ)/cc-pVTZ theory, of 0.87 kcal mol−1, results in a final prediction near 6.3 kcal mol−1 for the SiC2 barrier to linearity.

Sharing low-level functionality between software packages enables more rapid development of new capabilities and reduces the duplication of work among development groups. Using the component approach advocated by the Common Component Architecture Forum, we have designed a flexible interface for sharing integrals between quantum chemistry codes. Implementation of these interfaces has been undertaken within the Massively Parallel Quantum Chemistry package, exposing both the IntV3 and Cints/Libint integrals packages to component applications. Benchmark timings for Hartree-Fock calculations demonstrate that the overhead due to the added interface code varies significantly, from less than 1% for small molecules with large basis sets to nearly 10% for larger molecules with smaller basis sets. Correlated calculations and density functional approaches encounter less severe performance overheads of less than 5%. While these overheads are acceptable, additional performance losses occur when arbitrary implementation details, such as integral ordering within buffers, must be handled. Integral reordering is observed to add an additional overhead as large as 12%; hence, a common standard for such implementation details is desired for optimal performance.

Efficient design of hardware and software for large-scale parallel execution requires detailed understanding of the interactions between the application, computer, and network. The authors have developed a macro-scale simulator (SST/macro) that permits the coarse-grained study of distributed-memory applications. In the presented work, applications using the Message Passing Interface (MPI) are simulated; however, the simulator is designed to allow inclusion of other programming models. The simulator is driven from either a trace file or a skeleton application. Trace files can be either a standard format (Open Trace Format) or a more detailed custom format (DUMPI). The simulator architecture is modular, allowing it to easily be extended with additional network models, trace file formats, and more detailed processor models. This paper describes the design of the simulator, provides performance results, and presents studies showing how application performance is affected by machine characteristics.

Efficient design of hardware and software for large-scale parallel execution requires detailed understanding of the interactions between the application, computer, and network. The authors have developed a macro-scale simulator (SST/macro) that permits the coarse-grained study of distributed-memory applications. In the presented work, applications using the Message Passing Interface (MPI) are simulated; however, the simulator is designed to allow inclusion of other programming models. The simulator is driven from either a trace file or a skeleton application. Trace files can be either a standard format (Open Trace Format) or a more detailed custom format (DUMPI). The simulator architecture is modular, allowing it to easily be extended with additional network models, trace file formats, and more detailed processor models. This paper describes the design of the simulator, provides performance results, and presents studies showing how application performance is affected by machine characteristics.

The chemistry of the C 5 H 4 singlet potential energy surface was investigated at sophisticated levels of theory with a focus on the stability of pyramidane (tetracyclo[2.1.0.0 1,3 0 2,5 ]pentane or [3.3.3.3]fenestrane), a structure featuring a carbon atom at the apex of a square pyramid. Zero-point corrected relative energetics were predicted with both coupled cluster and density functional methodologies. Computations with both methodologies agree qualitatively with previous theoretical results, demonstrating that the pyramidane structure is a true minimum with substantial barriers to isomerization. At the CCSD(T)/TZ2P level a relative energy of 24 kcal/mol was predicted for the transition state to tricyclo[2.1.0.0 2,5 ]pent-3-ylidene, the lowest barrier to isomerization of pyramidane. The transition state to bicyclo[2.1.0]pent-2-ene-5-ylidene, the other transition state known to lead directly to the pyramidane structure, was found to lie 33 kcal/mol above pyramidane. Relative energies are also provided for several lower-lying C 5 H 4 isomers, including isomers incorporating linear carbon chains.