Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “cluster computing”

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 1,153 records · Page 64

Interconnected Cavernous Structure of Bacterial Fruiting Bodies

The formation of spore-filled fruiting bodies by myxobacteria is a fascinating case of multicelular self-organization by bacteria. The organization of Myxococcus xanthus into fruiting bodies has long been studied not only as an important example of collective motion of bacteria, but also as a simplified model for developmental morphogenesis. Sporulation within the nascent fruiting body requires signaling between moving cells in order that the rod-shaped self-propelled cells differentiate into spores at the appropriate time. Probing the three-dimensional structure of myxobacteria fruiting bodies has previously presented a challenge due to Imitations at different imaging methods. A new technique using Infrared Optical Coherence Tomography (OCT) revealed previously unknown details of the Internal structure of M. xanthus fruiting bodies consisting of interconnected pockets of relative nigh and low spore density regions. Here, to make sense of the experimentally observed structure, modeling and computer simulations were used to test a hypothesized mechanism that could produce high density pockets of spores. The mechanism consists of self-propelled cells aligning with each other and signaling by end-to-end contact to coordinate the process of differentiation resulting in a pattern of clusters observed in the experiment. The Integration of novel OCT experimental techniques with computational simulations can provide new insight Into the mechanisms that can give rise to the pattern formation seen In other biological systems such as dlctyostelids, social amoeba known to form multicellular aggregates observed as slugs under starvation conditions.

59 BASIC BIOLOGICAL SCIENCES↗

Accelerating DCA++ (Dynamical Cluster Approximation) Scientific Application on the Summit supercomputer

Optimizing scientific applications on today’s accelerator-based high performance computing systems can be challenging, especially when multiple GPUs and CPUs with heterogeneous memories and persistent non-volatile memories are present. An example is Summit, an accelerator-based system at the Oak Ridge Leadership Computing Facility (OLCF) that is rated as the world’s fastest supercomputer to-date. New strategies are thus needed to expose the parallelism in legacy applications, while being amenable to efficient mapping to the underlying architecture.In this paper we discuss our experiences and strategies to port a scientific application, DCA++, to Summit. DCA++ is a highperformance research application that solves quantum manybody problems with a cutting edge quantum cluster algorithm, the dynamical cluster approximation.Our strategies aim to synergize the strengths of the different programming models in the code. These include: (a) streamlining the interactions between the CPU threads and the GPUs, (b) implementing computing kernels on the GPUs and decreasing CPU-GPU memory transfers, (c) allowing asynchronous GPU communications, and (d) increasing compute intensity by combining linear algebraic operations.Full-scale production runs using all 4600 Summit nodes attained a peak performance of 73.5 PFLOPS with a mixed precision implementation.We observed a perfect strong and weak scaling for the quantum Monte Carlo solver in DCA++, while encountering about 2× input/output (I/O) and MPI communication overhead on the time-to-solution for the full machine run. Our hardware agnostic optimizations are designed to alleviate the communication and I/O challenges observed, while improving the compute intensity and obtaining optimal performance on a complex, hybrid architecture like Summit.

Balduzzi, Giovanni↗

Trajectory analysis via a geometric feature space approach

This study aimed to organize a body of trajectories in order to identify, search for and classify both common and uncommon behaviors among objects such as aircraft and ships. Existing comparison functions such as the Fréchet distance are computationally expensive and yield counterintuitive results in some cases. We propose an approach using feature vectors whose components represent succinctly the salient information in trajectories. These features incorporate basic information such as the total distance traveled and the distance between start/stop points as well as geometric features related to the properties of the convex hull, trajectory curvature and general distance geometry. Additionally, these features can generally be mapped easily to behaviors of interest to humans who are searching large databases. Most of these geometric features are invariant under rigid transformation. Furthermore, we demonstrate the use of different subsets of these features to identify trajectories similar to an exemplar, cluster a database of several hundred thousand trajectories and identify outliers.

97 MATHEMATICS AND COMPUTING↗

Molecular growth upon ionization of van der Waals clusters containing HCCH and HCN is a pathway to prebiotic molecules

The growth mechanisms of organic molecules in an ionizing environment such as the interstellar medium are not completely understood. Here we examine by means of ab initio molecular dynamics (AIMD) simulations and density functional theory (DFT) computations the possibility of bond formation and molecular growth upon ionization of van der Waals clusters of pure HCN clusters, and mixed clusters of HCN and HCCH, both of which are widespread in the interstellar medium. Ionization of van der Waals clusters can potentially lead to growth in low temperature and low-density environments. Our results show, that upon ionization of the pure HCN clusters, strongly bound stable structures are formed that contain N∙∙∙H bonds, and growth beyond pairwise HCN molecules is seen only in a small percentage of cases. In contrast, mixed clusters, where HCCH is preferentially ionized over HCN, can grow up to 3 or 4 units long with new carbon–carbon and carbon–nitrogen covalent bonds. Moreover, cyclic molecules formed, such as the radical cation of pyridine, which is a prebiotic molecule. The results presented here are significant as they provide a feasible pathway for molecular growth of small organic molecules containing both carbon and nitrogen in cold and relatively denser environments such as in dense molecular clouds but closer to the photo-dissociation regions, and protoplanetary disks. In the mechanism we propose, first, a neutral van der Waals cluster is formed. Once the cluster is formed it can undergo photoionization which leads to chemical reactivity without any reaction barrier.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Toward scalable quantum computations of atomic nuclei

We solve the nuclear two-body and three-body bound states via quantum simulations of pionless effective field theory on a lattice in position space. While the employed lattice remains small, the usage of local Hamiltonians including two- and three-body forces ensures that the number of Pauli terms scales linearly with increasing numbers of lattice sites. We use an adaptive ansatz grown from unitary coupled cluster theory to parametrize the ground states of the deuteron and 3 He, compute their corresponding energies, and analyze the scaling of the required computational resources. Our quantum simulations reproduce exact benchmarks for 2 H and 3 He within 100 keV, requiring at most 30 layers in the ansatz and thus resulting in modest circuit depths. Additionally, we find the number of shots required to reach a given precision scales linearly in the lattice size and more mildly in the system size. Furthermore, based on the agreement with exact benchmarks and mild scaling, we conclude that this can be an efficient, scalable approach for quantum computations of nuclear ground states, particularly to prepare initial states for quantum phase estimation or other filtering algorithms.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Data Intensive Computing on Amazon Web Services

The Geophysical Monitoring Program (GMP) has spent the past few years building up the capability to perform data intensive computing using what have been referred to as “big data” tools. These big data tools would be used against massive archives of seismic signals (>300 TB) to conduct research not previously possible. Examples of such tools include Hadoop (HDFS, MapReduce), HBase, Hive, Storm, Spark, Solr, and many more by the day. These tools are useful for performing data analytics on datasets that exceed the resources of traditional analytic approaches. To this end, a research big data cluster (“Cluster A”) was set up as a collaboration between GMP and Livermore Computing (LC).

97 MATHEMATICS AND COMPUTING↗

RCSB Protein Data Bank: visualizing groups of experimentally determined PDB structures alongside computed structure models of proteins

Recent advances in Artificial Intelligence and Machine Learning (e.g., AlphaFold, RosettaFold, and ESMFold) enable prediction of three-dimensional (3D) protein structures from amino acid sequences alone at accuracies comparable to lower-resolution experimental methods. These tools have been employed to predict structures across entire proteomes and the results of large-scale metagenomic sequence studies, yielding an exponential increase in available biomolecular 3D structural information. Given the enormous volume of this newly computed biostructure data, there is an urgent need for robust tools to manage, search, cluster, and visualize large collections of structures. Equally important is the capability to efficiently summarize and visualize metadata, biological/biochemical annotations, and structural features, particularly when working with vast numbers of protein structures of both experimental origin from the Protein Data Bank (PDB) and computationally-predicted models. Moreover, researchers require advanced visualization techniques that support interactive exploration of multiple sequences and structural alignments. This paper introduces a suite of tools provided on the RCSB PDB research-focused web portal RCSB. org, tailor-made for efficient management, search, organization, and visualization of this burgeoning corpus of 3D macromolecular structure data.

3D visualization↗

Diazotroph Community Characterization via a High-Throughput nifH Amplicon Sequencing and Analysis Pipeline

The dinitrogenase reductase gene ( nifH ) is the most widely established molecular marker for the study of nitrogen-fixing prokaryotes in nature. A large number of PCR primer sets have been developed for nifH amplification, and the effective deployment of these approaches should be guided by a rapid, easy-to-use analysis protocol. Bioinformatic analysis of marker gene sequences also requires considerable expertise. In this study, we advance the state of the art for nifH analysis by evaluating nifH primer set performance, developing an improved amplicon sequencing workflow, and implementing a user-friendly bioinformatics pipeline. Furthermore, the developed amplicon sequencing workflow is a three-stage PCR-based approach that uses established technologies for incorporating sample-specific barcode sequences and sequencing adapters. Based on our primer evaluation, we recommend the Ando primer set be used with a modified annealing temperature of 58°C, as this approach captured the largest diversity of nifH templates, including paralog cluster IV/V sequences. To improve nifH sequence analysis, we developed a computational pipeline which infers taxonomy and optionally filters out paralog sequences. In addition, we employed an empirical model to derive optimal operational taxonomic unit (OTU) cutoffs for the nifH gene at the species, genus, and family levels. A comprehensive workflow script named TaxADivA (TAXonomy Assignment and DIVersity Assessment) is provided to ease processing and analysis of nifH amplicons. Our approach is then validated through characterization of diazotroph communities across environmental gradients in beach sands impacted by the Deepwater Horizon oil spill in the Gulf of Mexico, in a peat moss-dominated wetland, and in various plant compartments of a sugarcane field. IMPORTANCE: Nitrogen availability often limits ecosystem productivity, and nitrogen fixation, exclusive to prokaryotes, comprises a major source of nitrogen input that sustains food webs. The nifH gene, which codes for the iron protein of the nitrogenase enzyme, is the most widely established molecular marker for the study of nitrogen-fixing microorganisms (diazotrophs) in nature. In this study, a flexible sequencing/analysis pipeline, named TaxADivA, was developed for nifH amplicons produced by Illumina paired-end sequencing, and it enables an inference of taxonomy, performs clustering, and produces output in formats that may be used by programs that facilitate data exploration and analysis. Diazotroph diversity and community composition are linked to ecosystem functioning, and our results advance the phylogenetic characterization of diazotroph communities by providing empirically derived nifH similarity cutoffs for species, genus, and family levels. Thus, the utility of our pipeline is validated for diazotroph communities in a variety of ecosystems, including contaminated beach sands, peatland ecosystems, living plant tissues, and rhizosphere soil.

59 BASIC BIOLOGICAL SCIENCES↗

COMPARISONS OF COSMOLOGICAL MAGNETOHYDRODYNAMIC GALAXY CLUSTER SIMULATIONS TO RADIO OBSERVATIONS

Radio observations of galaxy clusters show that there are {mu}G magnetic fields permeating the intracluster medium (ICM), but it is hard to accurately constrain the strength and structure of the magnetic fields without the help of advanced computer simulations. We present qualitative comparisons of synthetic Very Large Array observations of simulated galaxy clusters to radio observations of Faraday rotation measure (RM) and radio halos. The cluster formation is modeled using adaptive mesh refinement magnetohydrodynamic simulations with the assumption that the initial magnetic fields are injected into the ICM by active galactic nuclei (AGNs) at high redshift. In addition to simulated clusters in Xu et al., we present a new simulation with magnetic field injections from multiple AGNs. We find that the cluster with multiple injection sources is magnetized to a similar level as in previous simulations with a single AGN. The RM profiles from simulated clusters, both |RM| and the dispersion of RM ({sigma}{sub RM}), are consistent at a first order with the radial distribution from observations. The correlations between the {sigma}{sub RM} and X-ray surface brightness from simulations are in a broad agreement with the observations, although there is an indication that the simulated clusters could be slightly overdense and less magnetized with respect to those in the observed sample. In addition, the simulated radio halos agree with the observed correlations between the radio power versus the cluster X-ray luminosity and between the radio power versus the radio halo size. These studies show that the cluster-wide magnetic fields that originate from AGNs and are then amplified by the ICM turbulence match observations of magnetic fields in galaxy clusters.

79 ASTRONOMY AND ASTROPHYSICS↗

Probing electronic wave functions of sodium-doped clusters: Dyson orbitals, anisotropy parameters, and ionization cross-sections

We apply high-level ab initio methods to describe the electronic structure of small clusters of ammonia and dimethylether (DME) doped with sodium, which provide a model for solvated electrons. We investigate the effect of the solvent and cluster size on the electronic states. We consider both energies and properties, with a focus on the shape of the electronic wave function and the related experimental observables such as photoelectron angular distributions. The central quantity in modeling photoionization experiments is the Dyson orbital, which describes the difference between the initial N-electron and final (N-1)-electron states of a system. Dyson orbitals enter the expression of the photoelectron matrix element, which determines total and partial photoionization cross-sections. We compute Dyson orbitals for the Na(NH3)n and Na(DME)m clusters using correlated wave functions (obtained with equation-of-motion coupled-cluster model for electron attachment with single and double substitutions) and compare them with more approximate Hartree-Fock and Kohn-Sham orbitals. As a result, we also analyze the effect of correlation and basis sets on the shapes of Dyson orbitals and the experimental observables.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

The W40 region in the gould belt: An embedded cluster and H II region at the junction of filaments

We present a multiwavelength study of the W40 star-forming region using infrared (IR) observations in the UKIRT JHK bands, Spitzer Infrared Array Camera bands, and Herschel PACS bands, 2.12 μm H{sub 2} narrowband imaging, and radio continuum observations from GMRT (610 and 1280 MHz), in a field of view (FoV) of ∼34' × 40'. Archival Spitzer observations in conjunction with near-IR observations are used to identify 1162 Class II/III and 40 Class I sources in the FoV. The nearest-neighbor stellar surface density analysis shows that the majority of these young stellar objects (YSOs) constitute the embedded cluster centered on the high-mass source IRS 1A South. Some YSOs, predominantly the younger population, are distributed along and trace the filamentary structures at lower stellar surface density. The cluster radius is measured to be 0.44 pc—matching well with the extent of radio emission—with a peak density of 650 pc{sup –2}. The JHK data are used to map the extinction in the region, which is subsequently used to compute the cloud mass—126 M {sub ☉} and 71 M {sub ☉} for the central cluster and the northern IRS 5 region, respectively. H{sub 2} narrowband imaging shows significant emission, which prominently resembles fluorescent emission arising at the borders of dense regions. Radio continuum analysis shows that this region has a blister morphology, with the radio peak coinciding with a protostellar source. Free-free emission spectral energy distribution analysis is used to obtain physical parameters of the overall photoionized region and the IRS 5 sub-region. This multiwavelength scenario is suggestive of star formation having resulted from the merging of multiple filaments to form a hub. Star formation seems to have taken place in two successive epochs, with the first epoch traced by the central cluster and the high-mass star(s)—followed by a second epoch that is spreading into the filaments as uncovered by the Class I sources and even younger protostellar sources along the filaments. The IRS 5 H II region displays indications of swept-up material that has possibly led to the formation of protostars.

79 ASTRONOMY AND ASTROPHYSICS↗

Preparing NERSC users for Cori, a Cray XC40 system with Intel many integrated cores

The newest NERSC supercomputer Cori is a Cray XC40 system consisting of 2,388 Intel Xeon Haswell nodes and 9,688 Intel Xeon-Phi “Knights Landing” (KNL) nodes. Compared to the Xeon-based clusters NERSC users are familiar with, optimal performance on Cori requires consideration of KNL mode settings; process, thread, and memory affinity; fine-grain parallelization; vectorization; and use of the high-bandwidth MCDRAM memory. This paper describes our efforts preparing NERSC users for KNL through the NERSC Exascale Science Application Program, Web documentation, and user training. We discuss how we configured the Cori system for usability and productivity, addressing programming concerns, batch system configurations, and default KNL cluster and memory modes. Here, system usage data, job completion analysis, programming and running jobs issues, and a few successful user stories on KNL are presented.

97 MATHEMATICS AND COMPUTING↗

PULSATING B-TYPE STARS IN THE OPEN CLUSTER NGC 884: FREQUENCIES, MODE IDENTIFICATION, AND ASTEROSEISMOLOGY

Recent progress in the seismic interpretation of field β Cep stars has resulted in improvements of the physical description in the stellar structure and evolution model computations of massive stars. Further asteroseismic constraints can be obtained from studying ensembles of stars in a young open cluster, which all have similar age, distance, and chemical composition. We present an observational asteroseismology study based on the discovery of numerous multi-periodic and mono-periodic B stars in the open cluster NGC 884. We describe a thorough investigation of the pulsational properties of all B-type stars in the cluster. Overall, our detailed frequency analysis resulted in 115 detected frequencies in 65 stars. We found 36 mono-periodic, 16 bi-periodic, 10 tri-periodic, and 2 quadru-periodic stars and one star with nine independent frequencies. We also derived the amplitudes and phases of all detected frequencies in the U, B, V, and I filter, if available. We achieved unambiguous identifications of the mode degree for 12 of the detected frequencies in nine of the pulsators. Imposing the identified degrees and measured frequencies of the radial, dipole, and quadrupole modes of five pulsators led to a seismic cluster age estimate of log (age/yr) = 7.12-7.28 from a comparison with stellar models. Our study is a proof-of-concept for and illustrates the current status of ensemble asteroseismology of a young open cluster.

79 ASTRONOMY AND ASTROPHYSICS↗

Surface-supported cluster catalysis: Ensembles of metastable states run the show

It has recently been demonstrated that the dynamic behaviour of surface-supported nanocluster catalysts in realistic reaction conditions defies conventional models used in catalysis. This opens new doors in catalysis by giving more leverage in catalyst design, but also requires a major revision of the understanding of how dynamic heterogeneous catalytic interfaces operate, as well as of the computational approaches of catalyst modelling, and experimental methods of catalyst characterization. Key aspects of the new paradigm include the collective action of many catalyst states that form a statistical ensemble in reacting conditions, the catalytic activity and selectivity being driven by rare and metastable catalyst states, reaction thermodynamics and kinetics being controlled by different states of the catalyst, broken scaling relationships, non-Arrhenius behaviours, and catalyst dynamic restructuring being an essential part of the reaction mechanism. For computation, this complexity means the departure from the standard DFT calculations of reaction mechanisms on a single catalyst structure. For experiment, it calls for the development of operando characterization tools with the per-site resolution and the ability to find the minority sites that govern the catalytic activity. For catalyst design, the goal becomes the creation of the catalyst state (geometric and electronic) that might not be present in the as-prepared catalyst, but would develop in the reaction conditions and would have the desired activity then. While cluster catalysts are the most dramatic in their dynamic fluxionality, other amorphous interfaces also exhibit some of it, and thus are also subject to similar paradigm revision.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Computing the dipole polarizability of Ca 48 with increased precision

Here, we compute the electric dipole polarizability of 48 Ca with an increased precision by including more correlations than in previous studies. Employing the coupled-cluster method we go beyond single and double excitations and include leading-order three-particle-three-hole (3p-3h) excitations for the ground state, excited states, and the similarity-transformed operator. We study electromagnetic sum rules, such as the bremsstrahlung sum rule m 0 and the polarizability sum rule αD using interactions from chiral effective field theory. To gauge the quality of our coupled-cluster approximations we perform several benchmarks with the effective interaction hyperspherical harmonics approach in 4 He and with self consistent Green's function in 16 O. We compute the dipole polarizability of 48 Ca employing the chiral interaction N 2 LO sat and the 1.8/2.0 (EM). We find that the effect of 3p-3h excitations in the ground state is small for 1.8/2.0 (EM) but non-negligible for N 2 LO sat . The addition of these new correlations allows us to improve the precision of our 48 Ca calculations and reconcile the recently reported discrepancy between coupled-cluster results based on these interactions and the experimentally determined αD from proton inelastic scattering in 48 Ca. For the computation of electromagnetic and polarizability sum rules, the inclusion of leading-order 3p-3h excitations in the ground state is important, while it is less so for the excited states and the similarity-transformed dipole operator.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Linear-scaling quadruple excitations in local pair natural orbital coupled-cluster theory

Here, we present a fast, asymptotically linear-scaling implementation of the perturbative quadruples energy correction in coupled-cluster theory using local natural orbitals. Our work follows the domain-based local pair natural orbital (DLPNO) approach previously applied to lower levels of excitations in coupled-cluster theory. Our DLPNO-CCSDT(Q) algorithm uses converged doubles and triples amplitudes from a preceding DLPNO-CCSDT computation to compute the quadruples amplitude and energy in the quadruples natural orbital (QNO) basis. We demonstrate the compactness of the QNO space, showing that more than 95% of the (Q) correction can be recovered using relatively loose natural orbital cutoffs, compared to the tighter cutoffs used in pair and triples natural orbitals at lower levels of coupled-cluster theory. We also highlight the accuracy of our algorithm in the computation of relative energies, which yields deviations of sub-kJ mol −1 in relative energy compared to the canonical CCSDT(Q). Timings are conducted on a series of growing linear alkanes (up to 10 carbons and 608 basis functions) and water clusters (up to 49 water molecules and 2842 basis functions) to establish the asymptotic linear-scaling of our DLPNO-(Q) algorithm.

Auxiliary functions↗

Detecting Anomalies from End-to-End Internet Performance Measurements (PingER) Using Cluster Based Local Outlier Factor

PingER (Ping End-to-End Reporting) is a worldwide end-to-end Internet performance measurement framework. It was developed by the SLAC National Accelerator Laboratory, Stanford, USA and running from the last 20 years. It has more than 700 monitoring agents and remote sites which monitor the performance of Internet links around 170 countries of the world. At present, the size of the compressed PingER data set is about 60 GB comprising of 100,000 flat files. The data is publicly available for valuable Internet performance analyses. However, the data sets suffer from missing values and anomalies due to congestion, bottleneck links, queuing overflow, network software misconfiguration, hardware failure, cable cuts, and social upheavals. Therefore, the objective of this paper is to detect such performance drops or spikes labeled as anomalies or outliers for the PingER data set. In the proposed approach, the raw text files of the data set are transformed into a PingER dimensional model. The missing values are imputed using the k-NN algorithm. The data is partitioned into similar instances using the k-means clustering algorithm. Afterward, clustering is integrated with the Local Outlier Factor (LOF) using the Cluster Based Local Outlier Factor (CBLOF) algorithm to detect the anomalies or outliers from the PingER data. Lastly, anomalies are further analyzed to identify the time frame and location of the hosts generating the major percentage of the anomalies in the PingER data set ranging from 1998 to 2016.

97 MATHEMATICS AND COMPUTING↗

Clustering VoIP caller for SPIT identification

Abstract The number of unsolicited and advertisement telephony calls over traditional and Internet telephony has rapidly increased over recent few years. Every year, the telecommunication regulators, law enforcement agencies and telecommunication operators receive a very large number of complaints against these unsolicited, unwanted calls. These unwanted calls not only bring financial loss to the users of the telephony but also annoy them with unwanted ringing alerts. Therefore, it is important for the operators to block telephony spammers at the edge of the network so to gain trust of their customers. In this paper, we propose a novel spam detection system by incorporating different social network features for combating unwanted callers at the edge of the network. To this extent the reputation of each caller is computed by processing call detailed records of user using three social network features that are the frequency of the calls between caller and the callee, the duration between caller and the callee and the number of outgoing partners associated with the caller. Once the reputation of the caller is computed, the caller is then places in a spam and non‐spam clusters using unsupervised machine learning. The performance of the proposed approach is evaluated using a synthetic dataset generated by simulating the social behaviour of the spammers and the non‐spammers. The evaluation results reveal that the proposed approach is highly effective in blocking spammer with 2% false positive rate under a large number of spammers. Moreover, the proposed approach does not require any change in the underlying VoIP network architecture, and also does not introduce any additional signalling delay in a call set‐up phase. Copyright © 2016 John Wiley & Sons, Ltd.

Azad, Muhammad Ajmal↗