Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel processing”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 973 records · Page 54

Utilizing High Performance Computing for Chemistry: Parallel Computational Chemistry

Parallel hardware has become readily available to the computational chemistry research community. This perspective will review the current state of parallel computational chemistry software utilizing high-performance parallel computing platforms.Hardware and software trends and their effect on quantum chemistry methodologies, algorithms, and software development will also be discussed.

99 GENERAL AND MISCELLANEOUS↗

Silicon nanoparticle-functionalized fiberglass pads for sampling

We used wet treatment to immobilize luminescent silicon nanoparticles on industrial glass fibers to impart optical and chemical functions to the fiber. Carpets or pads consisting of thousands of fibers are processed in parallel, enhancing the sensitivity of detection and the sampled volume. Treated pads exhibit strong luminescence, characteristic of the luminescence of the particles; showing no shift, broadening, or reduction of quantum efficiency. We demonstrate that drawing material by the pad due to physical adsorption can be reversed. We also demonstrate that allylamine can be covalently attached by photoinduced irradiation reactions, which results in imprinting the amine emission spectrum, providing spectral recognition. The imprint accompanied with a blue-shifting of the luminescence spectrum of the probe, allowing examination of the effect of termination on the nanoparticle structure. The shift is found to be consistent with an increase in the bandgap of the Si nanoparticle and is consistent with Quantum Monte Carlo calculations. In addition to sampling, the nano probe pad has the potential to enable a variety of biomedical applications through subsequent attachment.

36 MATERIALS SCIENCE↗

Icing modelling in NSMB with chimera overset grids

In aerospace Engineering, the accurate simulation of ice accretion is a key element to increase flight safety and avoid accidents related to icing effects. The icing code developed in the NSMB solver is based on an Eulerian formulation for droplets tracking, an iterative Messinger model using a modified water runback scheme for ice thickness calculation and mesh deformation to track the ice/air interface through time. The whole process is parallelized with MPI and applied with chimera grids.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulated evolution of fluorophores for light emitting diodes

Organic light emitting diodes based on fluorophores with a propensity for thermally activated delayed fluorescence (TADF) are able to circumvent limitations imposed on device efficiency by spin statistics. Molecules with a propensity for TADF necessarily have two properties: a small gap between the lowest lying singlet and triplet excited states and a large transition dipole moment for fluorescence. In this work, we demonstrate the use of a genetic algorithm to search a region of chemical space for molecules with these properties. This algorithm is based on a flexible and intuitive representation of the molecule as a tree data structure, in which the nodes correspond to molecular fragments. Our implementation takes advantage of hybrid parallel graphics processing unit accelerated computer clusters to allow efficient sampling while retaining a reasonably accurate description of the electronic structure (in this case, CAM-B3LYP/6-31G{sup ∗∗}). In total, we have identified 3792 promising candidate fluorophores from a chemical space containing 1.26 × 10{sup 6} molecules. This required performing electronic structure calculations on only 7518 molecules, a small fraction of the full space. Several novel classes of molecules which show promise as fluorophores are presented.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A method for data handling numerical results in parallel OpenFOAM simulations

Parallel computational fluid dynamics simulations produce vast amount of numerical result data. This paper introduces a method for reducing the size of the data by replaying the interprocessor traffic. The results are recovered only in certain regions of interest configured by the user. A known test case is used for several mesh partitioning scenarios using the OpenFOAM toolkit{sup ®}[1]. The space savings obtained with classic algorithms remain constant for more than 60 Gb of floating point data. Our method is most efficient on large simulation meshes and is much better suited for compressing large scale simulation results than the regular algorithms.

42 ENGINEERING↗

Development of massively parallel quantum chemistry program SMASH

A massively parallel program for quantum chemistry calculations SMASH was released under the Apache License 2.0 in September 2014. The SMASH program is written in the Fortran90/95 language with MPI and OpenMP standards for parallelization. Frequently used routines, such as one- and two-electron integral calculations, are modularized to make program developments simple. The speed-up of the B3LYP energy calculation for (C{sub 150}H{sub 30}){sub 2} with the cc-pVDZ basis set (4500 basis functions) was 50,499 on 98,304 cores of the K computer.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Two-fluid and finite Larmor radius effects on helicity evolution in a plasma pinch

The evolution of magnetic energy, helicity, and hybrid helicity during nonlinear relaxation of a driven-damped plasma pinch is compared in visco-resistive magnetohydrodynamics and two-fluid models with and without the ion gyroviscous stress tensor. Magnetic energy and helicity are supplied via a boundary electric field which initially balances the resistive dissipation, and the plasma undergoes multiple relaxation events during the nonlinear evolution. The magnetic helicity is well conserved relative to the magnetic energy over each event, which is short compared with the global resistive diffusion time. The magnetic energy decreases by roughly 1.5% of its initial value over a relaxation event, while the magnetic helicity changes by at most 0.2% of the initial value. The hybrid helicity is dominated by magnetic helicity in low-β pinch conditions and is also well conserved. Differences of less than 1% between magnetic helicity and hybrid helicity are observed with two-fluid modeling and result from cross helicity evolution. The cross helicity is found to change appreciably due to the first-order finite Larmor radius effects which have not been included in contemporary relaxation theories. The plasma current evolves towards the flat parallel current state predicted by Taylor relaxation theory but does not achieve it. Plasma flow develops significant structure for two-fluid models, and the flow perpendicular to the magnetic field is much more substantial than the flow along it.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Solving global optimization problems on GPU cluster

The paper contains the results of investigation of a parallel global optimization algorithm combined with a dimension reduction scheme. This allows solving multidimensional problems by means of reducing to data-independent subproblems with smaller dimension solved in parallel. The new element implemented in the research consists in using several graphic accelerators at different computing nodes. The paper also includes results of solving problems of well-known multiextremal test class GKLS on Lobachevsky supercomputer using tens of thousands of GPU cores.

97 MATHEMATICS AND COMPUTING↗

Fully nonlinear δf gyrokinetics for scrape-off layer parallel transport

Edge plasmas present a few challenges for gyrokinetic simulations that are absent in tokamak cores. Among them are large fluctuation amplitudes and plasma-wall interactions in the open field line region. In this work, the widely used core turbulence code GENE, which employs a δf -splitting technique, is extended to simulate open systems with large electrostatic fluctuations. With inclusion and proper discretization of the parallel nonlinear term, it becomes equivalent to a full- f code and the δf -splitting causes no fundamental difficulty in handling large fluctuations. The loss of particles to the wall is accounted for by using a logical sheath boundary, which is implemented in the context of a finite-volume method. The extended GENE code is benchmarked for the well-established one-dimensional parallel transport problem in the scrape-off layer during edge-localized modes. The parallel heat flux deposited onto the divertor target is compared with previous simulation results and shows good agreement.

43 PARTICLE ACCELERATORS↗

Extended MHD modeling of tearing-driven magnetic relaxation

Discrete relaxation events in reversed-field pinch relevant configurations are investigated numerically with nonlinear extended magnetohydrodynamic modeling, including the Hall term in Ohm’s law and first-order ion finite Larmor radius effects. Our results show variability among relaxation events, where the Hall dynamo effect may help or impede the MHD dynamo effect in relaxing the parallel current density profile. The competitive behavior arises from multi-helicity conditions where the dominant magnetic fluctuation is relatively small. The resulting changes in parallel current density and parallel flow are aligned in the core, consistent with experimental observations. Analysis of simulation results also confirms that force density from fluctuation-induced Reynolds stress arises subsequent to the drive from fluctuation-induced Lorentz force density. Transport of momentum density is found to be dominated by the fluctuation-induced Maxwell stress over most of the cross section with viscous and gyroviscous contributions being large in the edge region. The findings resolve a discrepancy with respect to the relative orientation of current density and flow relaxation, which had not been realized or investigated in Ref. [King et. al. Phys. Plasmas 19, 055905 (2012)], where only the magnitude of flow relaxation is actually consistent with experimental results.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

The one-dimensional camelback potential in the parallel dipole line trap: Stability conditions and finite size effect

We recently demonstrated a magnetic parallel dipole line (PDL) system that serves as a unique diamagnetic trap with a fascinating one-dimensional camelback potential along its longitudinal axis. The system can be realized with a pair of transversely magnetized cylindrical magnets and a cylindrical graphite rod as the trapped object. In this paper, we present more detailed experimental and theoretical studies of the finite size effect of the rod and its impact on the stability and oscillation dynamics of the trap. We show that the camelback potential effect only occurs when the length of the PDL system is beyond certain critical length (L C ). The length of the trapped rod determines the “effective camelback potential” and is subject to maximum and minimum values for the trap to be stable. Both length and radius of the rod determine the damping dynamics or the quality factor of the oscillator. Finally, these characteristics are important for designing the PDL trap system for various sensing applications, for example, we demonstrated a PDL trap gas viscometer system through measurement of the oscillation damping time constant.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DIRECT INTEGRATION OF THE COLLISIONLESS BOLTZMANN EQUATION IN SIX-DIMENSIONAL PHASE SPACE: SELF-GRAVITATING SYSTEMS

We present a scheme for numerical simulations of collisionless self-gravitating systems which directly integrates the Vlasov-Poisson equations in six-dimensional phase space. Using the results from a suite of large-scale numerical simulations, we demonstrate that the present scheme can simulate collisionless self-gravitating systems properly. The integration scheme is based on the positive flux conservation method recently developed in plasma physics. We test the accuracy of our code by performing several test calculations, including the stability of King spheres, the gravitational instability, and the Landau damping. We show that the mass and the energy are accurately conserved for all the test cases we study. The results are in good agreement with linear theory predictions and/or analytic solutions. The distribution function keeps the property of positivity and remains non-oscillatory. The largest simulations are run on 64{sup 6} grids. The computation speed scales well with the number of processors, and thus our code performs efficiently on massively parallel supercomputers.

79 ASTRONOMY AND ASTROPHYSICS↗

AN ALGORITHM FOR RADIATION MAGNETOHYDRODYNAMICS BASED ON SOLVING THE TIME-DEPENDENT TRANSFER EQUATION

We describe a new algorithm for solving the coupled frequency-integrated transfer equation and the equations of magnetohydrodynamics in the regime that light-crossing time is only marginally shorter than dynamical timescales. The transfer equation is solved in the mixed frame, including velocity-dependent source terms accurate to O(v/c). An operator split approach is used to compute the specific intensity along discrete rays, with upwind monotonic interpolation used along each ray to update the transport terms, and implicit methods used to compute the scattering and absorption source terms. Conservative differencing is used for the transport terms, which ensures the specific intensity (as well as energy and momentum) are conserved along each ray to round-off error. The use of implicit methods for the source terms ensures the method is stable even if the source terms are very stiff. To couple the solution of the transfer equation to the MHD algorithms in the ATHENA code, we perform direct quadrature of the specific intensity over angles to compute the energy and momentum source terms. We present the results of a variety of tests of the method, such as calculating the structure of a non-LTE atmosphere, an advective diffusion test, linear wave convergence tests, and the well-known shadow test. We use new semi-analytic solutions for radiation modified shocks to demonstrate the ability of our algorithm to capture the effects of an anisotropic radiation field accurately. Since the method uses explicit differencing of the spatial operators, it shows excellent weak scaling on parallel computers.

79 ASTRONOMY AND ASTROPHYSICS↗

BALANCING THE LOAD: A VORONOI BASED SCHEME FOR PARALLEL COMPUTATIONS

One of the key issues when running a simulation on multiple CPUs is maintaining a proper load balance throughout the run and minimizing communications between CPUs. We propose a novel method of utilizing a Voronoi diagram to achieve a nearly perfect load balance without the need of any global redistributions of data. As a show case, we implement our method in RICH, a two-dimensional moving mesh hydrodynamical code, but it can be extended trivially to other codes in two or three dimensions. Our tests show that this method is indeed efficient and can be used in a large variety of existing hydrodynamical codes.

79 ASTRONOMY AND ASTROPHYSICS↗

Characterization and Improvement of RNA-Seq Precision in Quantitative Transcript Expression Profiling

Measurement precision determines the power of any analysis to reliably identify significant signals, such as in screens for differential expression, independent of whether the experimental design incorporates replicates or not. With the compilation of large scale RNA-Seq data sets with technical replicate samples, however, we can now, for the first time, perform a systematic analysis of the precision of expression level estimates from massively parallel sequencing technology. This then allows considerations for its improvement by computational or experimental means. Results: We report on a comprehensive study of target coverage and measurement precision, including their dependence on transcript expression levels, read depth and other parameters. In particular, an impressive target coverage of 84% of the estimated true transcript population could be achieved with 331 million 50 bp reads, with diminishing returns from longer read lengths and even less gains from increased sequencing depths. Most of the measurement power (75%) is spent on only 7% of the known transcriptome, however, making less strongly expressed transcripts harder to measure. Consequently, less than 30% of all transcripts could be quantified reliably with a relative error < 20%. Based on established tools, we then introduce a new approach for mapping and analyzing sequencing reads that yields substantially improved performance in gene expression profiling, increasing the number of transcripts that can reliably be quantified to over 40%. Extrapolations to higher sequencing depths highlight the need for efficient complementary steps. In discussion we outline possible experimental and computational strategies for further improvements in quantification precision.

59 BASIC BIOLOGICAL SCIENCES↗

3-D Simulations of earthquakes rupture jumps: 1. Homogeneous pre-stress conditions

SUMMARY Observational and modelling studies indicate that earthquake ruptures can jump between fault sections as large as ∼3 and ∼5 km for compressional and extensional offsets, respectively. Here, we compare characteristics of the rupture jump process on parallel but offset fault sections from traditional 3-D dynamic rupture simulations governed by slip weakening friction using the finite element code, FaultMod, to those from quasi-dynamic simulations governed by rate- and state-dependent friction (rate-state friction) using the code RSQSim. These simulations use spatially uniform initial stresses. For a variety of measures the rupture renucleation position on the offset fault, the rate-state friction and slip weakening friction models produce very similar results. The principal difference is the additional occurrence of delayed rupture jumps that arise from the time- and stress-dependent nucleation that is characteristic of rate-state friction. For immediate rupture jumps, models with slip weakening friction span greater offsets than those with rate-state friction. However the jump distances are nearly identical when delayed rupture jumps are included in the comparisons. We propose that delayed rupture jumps are the likely mechanism for adjacent large-earthquake pairs and clusters. Based on the similarity of renucleation positions with both dynamic and quasi-dynamic models, we conclude that the renucleation positions for rupture initiation on the receiver fault (separated by less than ∼3 km from the source fault) are primarily controlled by static stress changes induced by slip on the initiating fault. However, in light of the slightly greater maximum jump distances (>3 km) seen with the dynamic slip weakening friction model, dynamic stress changes from seismic waves play an increasingly important role as offset distances increase.

, RSQSim↗

Efficient On-demand Connection Management Mechanisms with PGAS Models on InfiniBand

In the last decade or so, clusters have observed a tremendous rise in popularity due to the excellent price to performance ratio. A variety of Interconnects have been proposed during this period, with InfiniBand leading the way due to its high performance and open standard. At the same time, multiple programming models have emerged in order to meet the requirements of various applications and their programming models. To support requirements of multiple programming models, InfiniBand provides multiple transport semantics, ranging from unreliable connectionless to reliable connected characteristics. Among them, the reliable connection (RC) semantics is being widely due to its high performance and support for novel features like Remote Direct Memory Acesss (RDMA), hardware atomics and Network Fault Tolerance. However, the pairwise connection oriented nature of the RC transport semantics limits its scalability and usage at the increasing processor counts. In this paper, we design and implement on-demand connection management approaches in the context of Partitioned Global Address Space (PGAS) programming models, which provided shared memory abstraction and one-sided communication semantics, leading to the development of multiple languages (UPC, X10, Chapel) and libraries (Global Arrays, MPI-RMA). Using Global Arrays as the research vehicle, we implement this approach with Aggregate Remote Memory Copy Interface (ARMCI), the runtime system of Global Arrays. We evaluate our approach, ARMCI-On Demand Connection Management (ARMCI-ODCM) using various microbenchmarks and benchmarks (LU Factorization, RandomAccess and Lennard Jones simulation) and application (Subsurface transport over multiple phases (STOMP)). With the performance evaluation for up to 4096 processors, we are able to have a multi-fold reduction in connection memory with a negligible degradation in performance. Using STOMP at 4096 processors, reduces the overall connection memory by 66 times with no performance degradation. To the best of our knowledge, this is the first design, implementation and evaluation of on-demand connection management with InfiniBand using PGAS models.

54 ENVIRONMENTAL SCIENCES↗