A Technique for Improving Performance of Global Collective Operations on Cluster of SMPs
Explore the source record for details and available documents.
SMP Cluster cluster computing parallel programming↗
SEARCH · Search NASA
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.
A community faulted-crust model using PYRAMID on cluster platforms
Explore the source record for details and available documents.
geophysics↗
Preliminary Evaluation of MapReduce for High-Performance Climate Data Analysis
MapReduce is an approach to high-performance analytics that may be useful to data intensive problems in climate research. It offers an analysis paradigm that uses clusters of computers and combines distributed storage of large data sets with parallel computation. We are particularly interested in the potential of MapReduce to speed up basic operations common to a wide range of analyses. In order to evaluate this potential, we are prototyping a series of canonical MapReduce operations over a test suite of observational and climate simulation datasets. Our initial focus has been on averaging operations over arbitrary spatial and temporal extents within Modern Era Retrospective- Analysis for Research and Applications (MERRA) data. Preliminary results suggest this approach can improve efficiencies within data intensive analytic workflows.
Duffy, Daniel Q.↗
Preliminary Analysis: Am-241 RHU/TEG Electric Power Source for Nanosatellites
The Februay 2013 Space Works Commercial report indicates a strong increase in nano/microsatellite (1-50 kg) launch demand globally in future years. Nanosatellites (NanoSats) are small spacecraft in the 1-10 kg range, which present a simple, low-cost option for developing quickly-deployable satellites. CubeSats, a special category of NanoSats, are even being considered for interplanetary missions. However, the small dimensions of CubeSats and the limited mass of the NanoSat class in general place limits of capability on their electrical power systems (especially where typical power sources such as solar panels are considered) and stored energy reserves; restricting the power budget and overall functionality. For example, leveraging NanoSat clusters for computationally intensive problems that are solved collectively becomes more challenging with power related restrictions on communication and data-processing. Further, interplanetary missions that would take NanoSats far from the sun, make the use of solar panels less effective as a power source as their required area would become quite large. To overcome these limitations, americium 241 (Am-241) has been suggested as a low power source option. The Idaho National Laboratory, Center for Space Nuclear Research reports that: (Production) requires small quantities of isotope - 62.5 g of Pu-238; 250 g Am- 241 (for 5 We); Am-241 is available at around 1 kg/yr commercially; Am-241 produces 59 kev gammas which are stopped readily by tungsten so the radiation field is very low. Whereby, an Am-241 source could be placed in among the instruments and the waste heat used to heat the platform; and amounts of isotope are so low that launch approval may be easier, especially with tungsten encapsulation. As further reported, Am-241 has a half-life that is approximately five times greater than that of Pu- 238 and it has been determined that the neutron yield of a 241-AmO(sub 2) source is approximately an order of magnitude lower than that of a 238-PuO(sub 2) source of equal mass and degree of (sup 16)O enrichment. Also it has been demonstrated that shielded heat sources fuelled by oxygen-enriched 238-PuO(sub 2) have masses that are up to 10 times greater than those fuelled by oxygenenriched 241-AmO(sub 2) with equivalent thermal power outputs and neutron dose rates at 1 m radii. For these reasons, Am-241 is well suited to missions that demand long duration electrical power output, such as deep spaceflight missions and similar missions that use radiation-hard electronics and instrumentation that are less susceptible to neutron radiation damage.
Robertson, Glen A.↗
Hadoop for High-Performance Climate Analytics: Use Cases and Lessons Learned
Scientific data services are a critical aspect of the NASA Center for Climate Simulations mission (NCCS). Hadoop, via MapReduce, provides an approach to high-performance analytics that is proving to be useful to data intensive problems in climate research. It offers an analysis paradigm that uses clusters of computers and combines distributed storage of large data sets with parallel computation. The NCCS is particularly interested in the potential of Hadoop to speed up basic operations common to a wide range of analyses. In order to evaluate this potential, we prototyped a series of canonical MapReduce operations over a test suite of observational and climate simulation datasets. The initial focus was on averaging operations over arbitrary spatial and temporal extents within Modern Era Retrospective- Analysis for Research and Applications (MERRA) data. After preliminary results suggested that this approach improves efficiencies within data intensive analytic workflows, we invested in building a cyber infrastructure resource for developing a new generation of climate data analysis capabilities using Hadoop. This resource is focused on reducing the time spent in the preparation of reanalysis data used in data-model inter-comparison, a long sought goal of the climate community. This paper summarizes the related use cases and lessons learned.
analytics↗
What Went Wrong: A Survey of Wildfire UAS Mishaps through Named Entity Recognition
Increasingly, unmanned aircraft systems (UAS) are being applied to wildfire incidents for tasks such as mapping, aerial ignition, and delivery. As a result, aviation incident reporting systems for wildfires are beginning to accumulate data related to UAS mishaps in wildfire response. In this research, we apply state-of-the-art natural language processing (NLP) techniques to develop a custom Named Entity Recognition (NER) model which extracts entities relevant to safety analysts. The custom NER model is built by fine-tuning an existing Bidirectional Encoder Representations from Transformers (BERT) model, resulting in a generalizable NER model that can extract engineering relevant entities including failure modes, causes, effects, control processes, and recommendations from failure-relevant text. This model performs passably, with a weighted average f1 score of 0.33 across entity types, indicating more labeled training data is needed. Extracted entities are used to form a Failure Modes and Effects Analysis (FMEA)-style survey of wildfire UAS mishaps reported using the SAFECOM system. Similar mishaps are manually clustered and reported as single rows within an FMEA. Foreach cluster, we compute frequency, severity, and overall riskin accordance with FAA standards. This methodology can beapplied as part of a broader safety management system totrack trends in mishaps (e.g., likelihood, severity) and discoverknowledge (e.g., causes, effects) that can be utilized to improvesafety outcomes and system performance.
Machine Learning↗
Ramdb: The NASA Raman Spectral Database (version 1.00).
Given that, in most instances, minimal sample preparation is required and due to its contactless instrument design, Raman spectroscopy is one of the most versatile vibrational spectroscopic techniques for the chemical analysis of environmental and biological specimens. The diversity of applications of Raman spectroscopy ranges anywhere from art [1] to planetary science missions [2]. The advancement in the use of Raman spectroscopy in Solar System missions, notably in post-mission sample return analysis, requires a spectral library holding the broad range of specimens that could be found in Solar System sources. For this purpose, we have initiated the development of a Raman spectral database (Ramdb) at NASA Ames Research Center. Currently, the database includes experimental and theoretical Raman spectra of PAHs [3, 4], as well as laboratory Raman spectra of amino acids, carbon allotropes, minerals, and analogs relevance to Earth Sciences [5], Exobiology [6], Planetary [7], and Astrochemistry [8] to name just a few examples. Ramdb can be found on the web at www.astrochemistry.org/ramdb, where raw and processed Raman spectra can be downloaded in CSV format. The laboratory Raman spectra are measured using a laser Raman spectrometer (JASCO NRS-5500-532QRI). The Raman instrument is equipped with three excitation lasers, with wavelengths of 405, 532, and 785 nm. A clean silicon substrate is used as the internal standard for wavenumber calibration. Powdered samples were prepared (microscopic >10 um, grounded microscopic < 10 um) on glass slides. Some raw data exhibited a background signal arising as a combination of laser-induced fluorescence from the sample. To correct this background, we developed a Python pipeline that uses open-source Python libraries. Ramdb provides both raw and processed (using Python pipeline) data, which includes tabulated Raman shift transitions and other measurement details. The theoretical Raman band positions of PAHs (pyrene monomers and tetramer clusters) were computed using density functional theory (DFT) with the help of the Gaussian 16 suite of programs [9]. In the near future, Ramdb will serve as a repository of Raman spectral data from Laboratory Astrophysics and Planetary Science experiments involving the irradiation of organic compounds under simulated space and planetary conditions. In addition, online and offline tools will be developed for utilising the database for comparison to the user’s sample.
N Punnakayathil↗
Improving DARHT High-Energy 20 MeV Gamma Ray Sources with the LAO Cluster
Explore the source record for details and available documents.
LAO cluster computer, hydro code, beam diagnostics↗
Architecture and method for a burst buffer using flash technology
A parallel supercomputing cluster includes compute nodes interconnected in a mesh of data links for executing an MPI job, and solid-state storage nodes each linked to a respective group of the compute nodes for receiving checkpoint data from the respective compute nodes, and magnetic disk storage linked to each of the solid-state storage nodes for asynchronous migration of the checkpoint data from the solid-state storage nodes to the magnetic disk storage. Each solid-state storage node presents a file system interface to the MPI job, and multiple MPI processes of the MPI job write the checkpoint data to a shared file in the solid-state storage in a strided fashion, and the solid-state storage node asynchronously migrates the checkpoint data from the shared file in the solid-state storage to the magnetic disk storage and writes the checkpoint data to the magnetic disk storage in a sequential fashion.
97 MATHEMATICS AND COMPUTING↗
Improving Block-level Efficiency with scsi-mq
Current generation solid-state storage devices are exposing a new bottlenecks in the SCSI and block layers of the Linux kernel, where IO throughput is limited by lock contention, inefficient interrupt handling, and poor memory locality. To address these limitations, the Linux kernel block layer underwent a major rewrite with the blk-mq project to move from a single request queue to a multi-queue model. The Linux SCSI subsystem rework to make use of this new model, known as scsi-mq, has been merged into the Linux kernel and work is underway for dm-multipath support in the upcoming Linux 4.0 kernel. These pieces were necessary to make use of the multi-queue block layer in a Lustre parallel filesystem with high availability requirements. We undertook adding support of the 3.18 kernel to Lustre with scsi-mq and dm-multipath patches to evaluate the potential of these efficiency improvements. In this paper we evaluate the block-level performance of scsi-mq with backing storage hardware representative of a HPC-targerted Lustre filesystem. Our findings show that SCSI write request latency is reduced by as much as 13.6%. Additionally, when profiling the CPU usage of our prototype Lustre filesystem, we found that CPU idle time increased by a factor of 7 with Linux 3.18 and blk-mq as compared to a standard 2.6.32 Linux kernel. Our findings demonstrate increased efficiency of the multi-queue block layer even with disk-based caching storage arrays used in existing parallel filesystems.
Cluster Computing↗
Segmentation of 9Cr Steel Samples Based on Composition and Mechanical Property
Explore the source record for details and available documents.
20 FOSSIL-FUELED POWER PLANTS↗
ExaGraph at Sandia: Graph Coloring Clustering and Partitioning for Exascale Computing.
Abstract not provided.
Boman, Erik Gunnar↗
Preface of the “Computational quantum chemistry and from small clusters to functional building blocks of novel nanomaterials symposium”
This symposium included papers with emphasis on Theoretical and Computational Quantum Chemistry and special focus on Materials Science perspectives.
36 MATERIALS SCIENCE↗
A new “gold standard”: Perturbative triples corrections in unitary coupled cluster theory and prospects for quantum computing
A major difficulty in quantum simulation is the adequate treatment of a large collection of entangled particles, synonymous with electron correlation in electronic structure theory, with coupled cluster (CC) theory being the leading framework for dealing with this problem. Augmenting computationally affordable low-rank approximations in CC theory with a perturbative account of higher-rank excitations is a tractable and effective way of accounting for the missing electron correlation in those approximations. This is perhaps best exemplified by the “gold standard” CCSD(T) method, which bolsters the baseline CCSD with the effects of triple excitations using considerations from many-body perturbation theory (MBPT). Despite this established success, such a synergy between MBPT and the unitary analog of CC theory (UCC) has not been explored. In this work, we propose a similar approach wherein converged UCCSD amplitudes are leveraged to evaluate energy corrections associated with triple excitations, leading to the UCCSD[T] method. In terms of quantum computing, this correction represents an entirely classical post-processing step that improves the energy estimate by accounting for triple excitation effects without necessitating new quantum algorithm developments or increasing demand for quantum resources. The rationale behind this choice is shown to be rigorous by studying the properties of finite-order UCC energy functionals, and our efforts do not support the addition of the fifth-order contributions as in the (T) correction. We assess the performance of these approaches on a collection of small molecules and demonstrate the benefits of harnessing the inherent synergy between MBPT and UCC theories.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Investigation of the cluster formation in lithium niobate crystals by computer modeling method
The processes occurring upon the formation of energetically equilibrium oxygen-octahedral clusters in the ferroelectric phase of a stoichiometric lithium niobate (LiNbO{sub 3}) crystal have been investigated by the computer modeling method within the semiclassical atomistic model. An energetically favorable cluster size (at which a structure similar to that of a congruent crystal is organized) is shown to exist. A stoichiometric cluster cannot exist because of the electroneutrality loss. The most energetically favorable cluster is that with a Li/Nb ratio of about 0.945, a value close to the lithium-to-niobium ratio for a congruent crystal.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
Ab Initio Calculations of Anharmonic Vibrational Spectroscopy for Hydrogen Fluoride (HF)n (n=3,4) and Mixed Hydrogen Fluoride/Water (HF)n(H20)n (n=1,2,4) Clusters
Anharmonic vibrational frequencies and intensities are computed for hydrogen fluoride clusters (HF)n with n=3,4 and mixed clusters of hydrogen fluoride with water (HF)n(H2O)n where n=1,2. For the (HF)4(H2O)4 complex, the vibrational spectra are calculated at the harmonic level, and anharmonic effects are estimated. Potential energy surfaces for these systems are obtained at the MP2/TZP level of electronic structure theory. Vibrational states are calculated from the potential surface points using the correlation-corrected vibrational self-consistent field (CC-VSCF) method. The method accounts for the anharmonicities and couplings between all vibrational modes and provides fairly accurate anharmonic vibrational spectra that can be directly compared with experimental results without a need for empirical scaling. For (HF)n, good agreement is found with experimental data. This agreement shows that the MP2 potential surfaces for these systems are reasonably reliable. The accuracy is best for the stiff intramolecular modes, which indicates the validity of MP2 in describing coupling between intramolecular and intermolecular degrees of freedom. For (HF)n(H2O)n experimental results are unavailable. The computed intramolecular frequencies show a strong dependence on cluster size. Intensity features are predicted for future experiments.
Chaban, Galina M.↗
Ab initio calculations of anharmonic vibrational spectroscopy for hydrogen fluoride (HF)n (n = 3, 4) and mixed hydrogen fluoride/water (HF)n(H2O)n (n = 1, 2, 4) clusters
Anharmonic vibrational frequencies and intensities are computed for hydrogen fluoride clusters (HF)n, with n = 3, 4 and mixed clusters of hydrogen fluoride with water (HF)n(H2O)n where n = 1, 2. For the (HF)4(H2O)4 complex, the vibrational spectra are calculated at the harmonic level, and anharmonic effects are estimated. Potential energy surfaces for these systems are obtained at the MP2/TZP level of electronic structure theory. Vibrational states are calculated from the potential surface points using the correlation-corrected vibrational self-consistent field method. The method accounts for the anharmonicities and couplings between all vibrational modes and provides fairly accurate anharmonic vibrational spectra that can be directly compared with experimental results without a need for empirical scaling. For (HF)n, good agreement is found with experimental data. This agreement shows that the Moller-Plesset (MP2) potential surfaces for these systems are reasonably reliable. The accuracy is best for the stiff intramolecular modes, which indicates the validity of MP2 in describing coupling between intramolecular and intermolecular degrees of freedom. For (HF)n(H2O)n experimental results are unavailable. The computed intramolecular frequencies show a strong dependence on cluster size. Intensity features are predicted for future experiments.
Water/chemistry↗