Alexander Moskovsky

Rigid-Body Molecular Dynamics of Fullerene-Based Nanocars on Metallic Surfaces

Methodical problems of coarse-grained-type molecular dynamics, namely, rigid-body molecular dynamics (RB MD), are studied by investigating the dynamics of nanosized molecular vehicles called nanocars that move on gold and silver surfaces. Specifically, we analyzed the role of thermostats and the effects of temperature, couplings, and correlations between rigid fragments of the nanocar molecule in extensive RB MD simulations. It is found that the use of the Nosé-Poincaré thermostat does not introduce systematic errors, but the time trajectories might be required to be limited to not accumulate large numerical integration errors. Correlations in the motion of different fragments of the molecules are also analyzed. Our theoretical computations also point to the importance of temperature, interfragment interactions, and interactions with surfaces and to the nature of the surface for understanding mechanisms of motion of single-molecule transporters.

DIATOMICS-IN-IONIC-SYSTEMS AND AB INITIO PREDICTIONS FOR THE STATIONARY POINTS ON POTENTIAL ENERGY SURFACES OF THE (HF)N CLUSTERS (N=3-6)

Predictions of the diatomics-in-ionic-systems model for the variety of stationary points on the potential energy surfaces of the hydrogen fluoride clusters (HF)n (3⩽n⩽6) are compared to the results of ab initio MP2/6-311+G(2d,2p) calculations as well as to the results of the polarizable mechanics model of Hodges et al. [J. Phys. Chem. A 102, 2455 (1998)]. The diatomics-in-ionic-systems scheme which relies on the balanced treatment of neutral and ionic contributions to the electronic properties of polyatomic species within the diatomics-in-molecules theory takes into account here the mixing of the FH and F−H+ electronic states. The corresponding mixing coefficient serves as a single principal adjustable parameter of the model, finally selected by the reference value of the binding energy of (HF)3. It is shown that structures and energies of the main cyclic isomers are in a good agreement with the best estimates of Quack and Suhm [Conceptual Perspectives in Quantum Chemistry (Kluwer, Dordrecht, 1997)]. Ever...

The Performance Characterization of the RSC PetaStream Module

The RSC PetaStream architecture is a massively parallel computer design based on Intel® Xeon® Phi manycore co-processors. Each RSC PetaStream module contains eight Intel Xeon Phi co-processors with PCI-express fabric and Infiniband interconnect for intermodule communication. This paper concentrates on the performance of a single RSC PetaStream module, evaluated with the help of low-level point-to-point MPI, library linear algebra, MAGMA and application-level classical molecular dynamics, GROMACS and LAMMPS codes tests. The Intel Xeon E5-2690 top bin CPU dual-socket system has been used for comparison. This early evaluation demonstrates that in general each Xeon Phi co-processor of RSC PetaStream delivers approximately the same performance as dual-socket Intel Xeon E5 system, with only a half energy-to-solution. Fine-grain parallelism of Intel Xeon Phi cores takes advantage of higher messages exchange rates on MPI level for communication of threads placed on different Xeon Phi chips.

On Quantum Chemistry Code Adaptation for RSC PetaStream Architecture

Molecular simulations with quantum chemistry methods consume a large portion of CPU cycles in modern high-performance computing centers. Evolution of modern processors and HPC architectures necessitates adaptation of software to new hardware generations. The present work concentrates on the optimization of the widely used GAMESS code to Intel Xeon Phi architecture and recently devised RSC PetaStream platform. Since improvement in parallelization is required, the most frequently used Hartree-Fock and DFT methods are explored for additional parallelization options. The Xeon Phi requires vectorization that is important for electron-repulsion integrals (ERI) calculations to achieve good performance.

An analysis of stationary points on the (HF)n potential surfaces (n≤6) predicted by the diatomics-in-ionic-systems model

Abstract The lowest energy stationary points on the potential energy surfaces of the hydrogen fluoride clusters (HF) n (4≤ n ≤6) predicted by the semiempirical diatomics-in-ionic-systems (DIIS) approach are verified by using ab initio MP2/aug-cc-pVDZ, MP2/6-311+G ∗∗ and hybrid B3LYP/6-311+G ∗∗ calculations. The results are also compared with the predictions of the polarizable mechanics model [Hodges et al., J. Phys. Chem. A, 102 (1998) 2455]. It is concluded that the DIIS scheme which is an inexpensive tool to scan large areas of potential energy surfaces may serve as a guide for the search of global and local minima points, with a reliability competitive to other approaches in the theory of intermolecular interactions.

Optical transitions in the light-harvesting complexes of bacterial photosynthetic centers

The geometric parameters of the model structure of the bacterial photosynthetic center, which consists of a light-harvesting antenna complex, a reaction center, lipid layers, and the solvation shells of water molecules, were optimized using quantum mechanics-molecular mechanics (QM/MM). The optical absorption spectra of the bacteriochlorophyll were evaluated in the quantum subsystem. It was shown that the shift of the absorption bands towards longer wavelengths as compared to the spectrum of the individual bacteriochlorophyll molecules is characteristic of the cluster of pigment molecules that are found in the antennas.

Abstract: Using Active Storage Concept for Seismic Data Processing

This poster presents an approach to distributed seismic data processing using Seismic Un*x Software and the Active Storage system based on TSim C++ template library and the Lustre file system. The study analyzes performance results of the developed system as well as the usability of Active Storage and Seismic Unix integration.

Modeling the Solvation Sites in Rare-Gas Matrices with the Simulated Annealing Monte Carlo Technique

The formation of solvation sites of rare-gas atoms (Rg = Ar, Kr) that trap the guest molecular systems is modeled using the simulated annealing Monte Carlo technique. As the guests, two molecular species are considered namely, the simplest hydrogen-bonded intermolecular complex (HF)2, and the radical SH in the excited electronic state. Both are important for investigations in the low-temperature matrices. The realistic potential energy surfaces for the simulations are constructed on the basis of spectroscopically parametrized functions for the constituting fragments and empirical solvent−solvent interaction potentials. It is shown that both examples give distinctively different pictures of the solvation sites. For the (HF)2@Arn clusters three types of solvation shells can be distinguished which correlate with the structures of pure argon clusters as well as with the structure in the bulk. For the SH@Rgn clusters no clear regularities are seen, and larger amounts of solvent atoms are needed to form the tra...

An efficient MPI/openMP parallelization of the Hartree-Fock method for the second generation of Intel ® Xeon Phi ™ processor

Modern OpenMP threading techniques are used to convert the MPI-only Hartree-Fock code in the GAMESS program to a hybrid MPI/OpenMP algorithm. Two separate implementations that differ by the sharing or replication of key data structures among threads are considered, density and Fock matrices. All implementations are benchmarked on a super-computer of 3,000 Intel® Xeon Phi™ processors. With 64 cores per processor, scaling numbers are reported on up to 192,000 cores. The hybrid MPI/OpenMP implementation reduces the memory footprint by approximately 200 times compared to the legacy code. The MPI/OpenMP code was shown to run up to six times faster than the original for a range of molecular system sizes.

The Energy Consumption Analysis for the Multispectral Infrared Satellite Images Processing Algorithm

This paper includes the energy consumption analysis of the testing mini-application that implements night time infrared remote sensing algorithm Nightfire. On this stage of our project computational nodes with Intel Xeon E5 and Intel Xeon Phi processors were tested.