Mutsumi Aoyagi

NONADIABATIC BENDING DISSOCIATION IN 16 VALENCE ELECTRON SYSTEM OCS

The speed, angular, and alignment distributions of S(1D2) atoms from the ultraviolet photodissociation of OCS have been measured by a photofragment imaging technique. From the excitation wavelength dependence of the scattering distribution of S(1D2), the excited states accessed by photoabsorption were assigned to the A′ Renner–Teller component of the  1Δ and the A″(1Σ−) states. It was found that the dissociation from the A′ state gives rise to high- and low-speed fragments, while the A″ state only provides the high-speed fragment. In order to elucidate the dissociation dynamics, in particular the bimodal speed distribution of S atoms, two-dimensional potential energy surfaces of OCS were calculated for the C–S stretch and bending coordinates by ab initio molecular orbital (MO) configuration interaction (CI) method. Conical intersections of 1Δ and 1Σ− with 1Π were found as adiabatic dissociation pathways. Wave packet calculations on these adiabatic surfaces, however, did not reproduce the low-speed compone...

Performance prediction of large-scale parallell system and application using macro-level simulation

To predict application performance on an HPC system is an important technology for designing the computing system and developing applications. However, accurate prediction is a challenge, particularly, in the case of a future coming system with higher performance. In this paper, we present a new method for predicting application performance on HPC systems. This method combines modeling of sequential performance on a single processor and macro-level simulations of applications for parallel performance on the entire system. In the simulation, the execution flow is traced but kernel computations are omitted for reducing the execution time. Validation on a real terascale system showed that the predicted and measured performance agreed within 10% to 20 %. We employed the method in designing a hypothetical petascale system of 32768 SIMD-extended processor cores. For predicting application performance on the petascale system, the macro-level simulation required several hours.

Analyzing and mitigating the impact of manufacturing variability in power-constrained supercomputing

A key challenge in next-generation supercomputing is to effectively schedule limited power resources. Modern processors suffer from increasingly large power variations due to the chip manufacturing process. These variations lead to power inhomogeneity in current systems and manifest into performance inhomogeneity in power constrained environments, drastically limiting supercomputing performance. We present a first-of-its-kind study on manufacturing variability on four production HPC systems spanning four microarchitectures, analyze its impact on HPC applications, and propose a novel variation-aware power budgeting scheme to maximize effective application performance. Our low-cost and scalable budgeting algorithm strives to achieve performance homogeneity under a power constraint by deriving application-specific, module-level power allocations. Experimental results using a 1,920 socket system show up to 5.4X speedup, with an average speedup of 1.8X across all benchmarks when compared to a variation-unaware power allocation scheme.

Millimeter-wave spectroscopy of the internal-rotation band of the He-HCN complex and the intermolecular potential energy surface

Millimeter-wave absorption spectroscopy combined with a pulsed-jet expansion technique was applied to measure the internal-rotation band of He–HCN in the frequency region of 95–125 GHz. In total 13 rovibrational lines, split into nitrogen nuclear hyperfine structure, were observed for the fundamental internal-rotation band, j=1−0. The observed transition frequencies were analyzed including their hyperfine splitting to yield an intermolecular potential energy surface, as improved from the one given by a coupled-cluster single double (triple) ab initio calculation. The surface obtained has a global minimum in the linear configuration (He⋅⋅⋅H–C–N) with a well depth of 30.2 cm−1, and a saddle point located in the antilinear configuration (H–C–N⋅⋅⋅He) which is higher by 8.91 cm−1 in energy than the global minimum. The distance Rm between the He atom and the center of mass of HCN along the minimum energy path shows a strong angular dependence; Rm is 4.169 and 4.040 A in the linear and antilinear forms, respecti...

Theoretical study of the potential energy surfaces and dynamics of CaNC/CaCN

Potential energy surfaces for the ground and two low-lying electronically excited states of CaNC/CaCN, are calculated using the ab initio molecular orbital (MO) configuration interaction (CI) method. The absorption and emission spectra of the system are computed by performing time-dependent quantum dynamical calculations on these surfaces. The most stable geometries for the two lowest lying 12Σ+ and 12Π electronic states correspond to the calcium isocyanide (CaNC) structure. These two states are characterized by ionic bonding and the potential energy curves along the bending coordinate are relatively isotropic. The result of our wave packet dynamics shows that the characteristics of the experimental spectra observed by the laser-induced fluorescence spectroscopy can be explained by the Renner–Teller splitting.

CHEMICAL REACTIONS IN THE O(1D) + HCl SYSTEM III.

Using the accurate global potential energy surfaces for the 11A′, 11A′′, and 21A′ states reported in the previous sister Paper I, detailed quantum dynamics calculations are performed for these three adiabatic surfaces separately for J = 0 (J: total angular momentum quantum number). Overall reaction probabilities for O + HCl → OH + Cl and H + ClO, the branching ratio between the two reactions, effects of the initial rovibrational excitation, and product rovibrational distributions are evaluated in the total energy region Etot ≤ 0.9 eV. Significant contributions to the overall reaction dynamics are found from the two excited 11A′′ and 21A′ potential energy surfaces, clearly indicating the insufficiency of the dynamics only on the ground 11A′ surface. The detailed dynamics on the excited surfaces are reported in the third paper of this series.

CHEMICAL REACTIONS IN THE O(1D) + HCl SYSTEM II.: DYNAMICS ON THE GROUND 11A′ STATE AND CONTRIBUTIONS OF THE EXCITED (11A″ AND 21A′) STATES

Using the accurate global potential energy surfaces for the 11A′, 11A′′, and 21A′ states reported in the previous sister Paper I, detailed quantum dynamics calculations are performed for these three adiabatic surfaces separately for J = 0 (J: total angular momentum quantum number). Overall reaction probabilities for O + HCl → OH + Cl and H + ClO, the branching ratio between the two reactions, effects of the initial rovibrational excitation, and product rovibrational distributions are evaluated in the total energy region Etot ≤ 0.9 eV. Significant contributions to the overall reaction dynamics are found from the two excited 11A′′ and 21A′ potential energy surfaces, clearly indicating the insufficiency of the dynamics only on the ground 11A′ surface. The detailed dynamics on the excited surfaces are reported in the third paper of this series.

CHEMICAL REACTIONS IN THE O(1D) + HCl SYSTEM III.: QUANTUM DYNAMICS ON THE EXCITED (11A″ AND 21A′) POTENTIAL ENERGY SURFACES

Using the accurate global potential energy surfaces for the 11A′′ and 21A′ states reported in the previous sister Paper I, detailed quantum dynamics calculations are performed for these adiabatic surfaces separately for J = 0 (J: total angular momentum quantum number). In addition to the significant overall contributions of these states to the title reactions reported in the second Paper II of this series, quantum dynamics on these excited potential energy surfaces (PES) are clarified in terms of the PES topographies, which are quite different from that of the ground PES. The reaction mechanisms are found to be strongly selective and nicely explained as vibrationally nonadiabatic transitions in the vicinity of potential ridge.

Power Consumption Evaluation of an MHD Simulation with CPU Power Capping

Recently to achieve the Exa-flops next generation computer system, the power consumption becomes the important issue. On the other hand, the power consumption character of application program is not so considered now. In this study we examine the power character of our Magneto hydrodynamic (MHD) simulation code for the global magnetosphere to evaluate the power consumption behavior of the simulation code under the CPU power capping on the parallel computer system. As a result, it is confirmed that there are different power consumption parts in the MHD simulation code, which the execution performance decreases or does not change under the CPU power capping. This indicates the capability of performance optimization with the power capping.

Theoretical study of the potential energy surfaces and bound states of HCP

Global, ab initio potential energy surfaces for HCP in its ground 1 1Σ+ (1 1A′) and low-lying excited 1 1A″, 2 1A′, and 1 1Δ(2 1A″) electronic states are determined. The multireference configuration interaction method at the double zeta with polarization basis set level is used, although some calculations augmented with diffuse functions are also discussed. Numerous quantum mechanical rovibrational states are then obtained for these surfaces, with emphasis on those corresponding to excited electronic state levels which have not been studied theoretically before. The results agree reasonably well with available experimental data for the 1 1A″ state. Furthermore, the presence of certain local minima on the 1 1A″ and 2 1A′ surfaces leads to one new series of levels on the 1 1A″ surface, and two new series on the 2 1A′ surface.