Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel, scaling”

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 1,801 records · Page 100

Galactos: computing the anisotropic 3-point correlation function for 2 billion galaxies

The nature of dark energy and the complete theory of gravity are two central questions currently facing cosmology. A vital tool for addressing them is the 3-point correlation function (3PCF), which probes deviations from a spatially random distribution of galaxies. However, the 3PCF's formidable computational expense has prevented its application to astronomical surveys comprising millions to billions of galaxies. We present Galactos, a high-performance implementation of a novel, O( N 2 ) algorithm that uses a load-balanced k-d tree and spherical harmonic expansions to compute the anisotropic 3PCF. Our implementation is optimized for the Intel Xeon Phi architecture, exploiting SIMD parallelism, instruction and thread concurrency, and significant L1 and L2 cache reuse, reaching 39% of peak performance on a single node. Galactos scales to the full Cori system, achieving 9.8 PF (peak) and 5.06 PF (sustained) across 9636 nodes, making the 3PCF easily computable for all galaxies in the observable universe.

79 ASTRONOMY AND ASTROPHYSICS↗

CORRLA-RS

The CORRLA-RS package provides a suite of statistical methods for sampling multidimensional distributions and to conduct sensitivity and correlation analysis of large scale data in the Rust programming language. The software provides a unique solution to multidimensional constrained sampling problems utilizing a combination of parallelized Markov Chain Monte Carlo methods and traditional rejection sampling. The sensitivity and correlation analysis methods are backed by a high performance randomized singular value decomposition implementation which enables datasets larger than the random access memory (RAM) size to be analyzed. Additionally, CORRLA-RS implements the active subspace identification method using a KD-Tree and the randomized singular value decomposition acting in concert.

Gurecky, William [Oak Ridge National Laboratory (O↗

Numerical eigen-spectrum slicing, accurate orthogonal eigen-basis, and mixed-precision eigenvalue refinement using OpenMP data-dependent tasks and accelerator offload

Performing a variety of numerical computations efficiently and, at the same time, in a portable fashion requires both an overarching design followed by a number of implementation strategies. All of these are exemplified below as we present transitioning the PLASMA numerical library from relying on dependence-driven large tasks to achieving utilization of fine grain tasking and offload to hardware accelerators while keeping its core dependence sets: OpenMP source code pragmas and runtime for most system-level functionality and basic low-level numerical kernels provided directly by hardware vendors or open source projects with vendor contributions. We also present new algorithmic methods and their efficient parallel implementations including fine grained tasking for eigen-spectrum slicing and offload for mixed-precision eigenvalue refinement. We provide performance, scaling, and numerical results showing sizable gains over the available solutions from either the open source and vendor-provided packages.

Luszczek, Piotr↗

Scintillating Bubble Chambers for Rare Event Searches

The Scintillating Bubble Chamber (SBC) collaboration is developing liquid-noble bubble chambers for the detection of sub-keV nuclear recoils. These detectors benefit from the electron recoil rejection inherent in moderately-superheated bubble chambers with the addition of energy reconstruction provided from the scintillation signal. The ability to measure low-energy nuclear recoils allows the search for GeV-scale dark matter and the measurement of coherent elastic neutrino-nucleus scattering on argon from MeV-scale reactor antineutrinos. The first physics-scale detector, SBC-LAr10, is in the commissioning phase at Fermilab, where extensive engineering and calibration studies will be performed. In parallel, a functionally identical low-background version, SBC-SNOLAB, is being built for a dark matter search underground at SNOLAB. SBC-SNOLAB, with a 10 kg-yr exposure, will have sensitivity to a dark matter–nucleon cross section of 2×10 –42 cm 2 at 1 GeV/c 2 dark matter mass, and future detectors could reach the boundary of the argon neutrino fog with a tonne-yr exposure. In addition, the deployment of an SBC detector at a nuclear reactor could enable neutrino physics investigations including measurements of the weak mixing angle and searches for sterile neutrinos, the neutrino magnetic moment, and the light Z’ gauge boson.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Performance of Replica-Exchange Wang-Landau Sampling for the 2D Ising Model: A Brief Survey

We report a brief performance study of the replica-exchange Wang-Landau algorithm, a recently proposed parallel realization of Wang-Landau sampling, using the 2D Ising model as a test case. The simulation time is found to scale inversely with the square root of the number of subwindows (and thus number of processors) used to span the global parameter space. We also investigate the time profiles for random walkers in dierent subwindows to complete iterations, which will aid the development of and adaptive load-balancing scheme.

Zhao, Yiwei↗

Operating System Support for Fine-Grained Task Migration

With the end of clock scaling and the limited power budget available (20-30MW), future supercomputers will meet exascale performance primarily through a higher level of parallelism. Current operating (OS) and runtime (RT) systems are designed for the classical SMP model and based on the static and coarse-grained process/thread paradigm. They do not provide the required level of flexibility, especially within a single compute node, to meet the requirements imposed by exascale systems in terms of power/energy efficiency, resilience, managing concurrency and performance portability. The high level of concurrency poses new challenges specific to exascale systems that need to be addressed by novel solutions. In particular, requiring the user to manage billions of concurrent threads could easily result in poor data locality, clogged interconnection networks, unchecked propagation of soft errors, and lack of control over power/energy consumption. Equally important, managing such level of concurrency interferes with the user’s focus on the application and the algorithm. We envision that the computation will be encapsulated into fine-grained tasks that can be isolated and protected from the other tasks running in the system. Whenever a task needs to work on some data that is not stored on the local node, the OS/RT allows the task to be migrated to the node that owns the data. Each task is associated with a contained state (set or architectural registers, stack frame, running node, etc.) that describes the progress of the task and that should be moved together with the task’s code. We believe system support for task migration is a fundamental function that can ease the job of tackling several of the exascale challenges.

Gioiosa, Roberto↗

MAGNETIC BRAIDING AND QUASI-SEPARATRIX LAYERS

The squashing factor Q, a property of the magnetic field line mapping, has been suggested as an indicator for the formation of current sheets, and subsequently magnetic reconnection, in astrophysical plasmas. Here, we test this hypothesis for a particular class of braided magnetic fields which serve as a model for solar coronal loops. We explore the relationship between quasi-separatrix layers (QSLs), that is, layer-like structures with high Q value, electric currents, and integrated parallel currents; the latter being a quantity closely related to the reconnection rate. It is found that as the degree of braiding of the magnetic field is increased, the maximum values of Q increase exponentially. At the same time, the distribution of Q becomes increasingly filamentary, with the width of the high-Q layers exponentially decreasing. This is accompanied by an increase in the number of layers so that as the field is increasingly braided the volume becomes occupied by a myriad of thin QSLs. QSLs are not found to be good predictors of current features in this class of braided fields. Indeed, despite the presence of multiple QSLs, the current associated with the field remains smooth and large scale under ideal relaxation; the field dynamically adjusts to a smooth equilibrium. Regions of high Q are found to be better related to regions of high integrated parallel current than to actual current sheets.

79 ASTRONOMY AND ASTROPHYSICS↗

Large-scale real-time signal processing in physics experiments: the ALICE TPC FPGA pipeline

For LHC Run 3, the ALICE Time Projection Chamber was upgraded to operate in continuous readout mode. Interaction rates of up to 50 kHz in Pb-Pb collisions require real-time processing of more than 3 TB s -1 of raw detector data. This requirement is met by a custom FPGA-based processing pipeline that performs the complete front-end data treatment fully in-stream, including common-mode correction, pedestal subtraction, ion-tail filtering, zero suppression, and dense data packing. A central element of the design is a highly parallel common-mode correction algorithm operating directly on the streaming data. It robustly identifies signal-free readout channels on a time-bin basis and applies pad-dependent scaling to compensate for local variations in capacitive coupling in the GEM readout. In combination with pedestal subtraction and ion-tail filtering, this enables accurate baseline restoration under extreme high-occupancy conditions, preventing signal loss while efficiently suppressing noise prior to zero suppression. The pipeline operates continuously at the full detector bandwidth and reduces the raw input rate of approximately 3 TB s -1 to about 900 GBps for Pb-Pb collisions at the target interaction rate. Overall, it represents a large-scale FPGA-based real-time signal-processing implementation for high-energy physics detector readout.

Digital signal processing (DSP)↗

Experimental Investigation of Low-Frequency Distributed Acoustic Sensor Responses to Two Parallel Propagating Fractures

Low-frequency distributed acoustic sensing (LF-DAS) is a diagnostic tool for hydraulic fracture propagation with far-field monitoring using fiber optic sensors. LF-DAS senses strain rate variation caused by stress field change due to fracture propagation. Fiber optic sensors are installed in the monitoring wells in the vicinity of a fractured well. From the strain responses, fracture propagation can be evaluated. To understand subsurface conditions with multiple propagating fractures, a laboratory-scale hydraulic fracture experiment was performed simulating the LF-DAS response to fracture propagation with embedded distributed optical fiber strain sensors under these conditions. The experiment was performed using a transparent cube of epoxy with two parallel radial initial flaws centered in the cube. Fluid was injected into the sample to generate fractures along the initial flaws. The experiment used distributed high-definition fiber optic strain sensors with tight spatial resolutions. The sensors were embedded at two different locations on opposite sides of the initial flaws, serving as observation/monitoring locations. We also employed finite element modeling to numerically solve the linear elastic equations of equilibrium continuity and stress–strain relationships. The measured strains from the experiment were compared to simulation results from the finite element model. The experimentally derived strain and strain-rate waterfall plots from this study show the responses to both fractures propagating, while the fracture at the lower position took most of the fluid during the experiment. Interestingly, a fracture first began propagating from the upper flaw of the two flaws, but once the lower fracture was initiated, it grew much faster than the upper fracture. Both fibers were intercepted by the lower fracture, further verifying the strain signature as a fracture is approaching and intersecting an offset fiber.

Chemistry↗

PuLP/XtraPuLP : Partitioning Tools for Extreme-Scale Graphs

PuLP/XtraPulp is software for partitioning graphs from several real-world problems. Graphs occur in several places in real world from road networks, social networks and scientific simulations. For efficient parallel processing these graphs have to be partitioned (split) with respect to metrics such as computation and communication costs. Our software allows such partitioning for massive graphs.

Slota, GeorgeM↗

ArrayBridge: Interweaving declarative array processing with high-performance computing

Scientists are increasingly turning to datacenter-scale computers to produce and analyze massive arrays. Despite decades of database research that extols the virtues of declarative query processing, scientists still write, debug and parallelize imperative HPC kernels even for the most mundane queries. This impedance mismatch has been partly attributed to the cumbersome data loading process; in response, the database community has proposed in situ mechanisms to access data in scientific file formats. Scientists, however, desire more than a passive access method that reads arrays from files. This paper describes ArrayBridge, a bi-directional array view mechanism for scientific file formats, that aims to make declarative array manipulations interoperable with imperative file-centric analyses. Our prototype implementation of ArrayBridge uses HDF5 as the underlying array storage library and seamlessly integrates into the SciDB open-source array database system. In addition to fast querying over external array objects, ArrayBridge produces arrays in the HDF5 file format just as easily as it can read from it. ArrayBridge also supports time travel queries from imperative kernels through the unmodified HDF5 API, and automatically deduplicates between array versions for space efficiency. Our extensive performance evaluation in NERSC, a large-scale scientific computing facility, shows that ArrayBridge exhibits statistically indistinguishable performance and I/O scalability to the native SciDB storage engine.

97 MATHEMATICS AND COMPUTING↗

Impact of High Performance Computing on the Detailed Simulation of Subsurface Flow and Contaminant Transport Using eSTOMP - 16341

The ability to model subsurface processes at the field-scale is essential to site closure and remedial action decisions. Field-scale models often require high-performance computing to meet the computational demands associated with complex systems, model calibration, and uncertainty quantification. eSTOMP, a parallel version of STOMP, is an analytical tool that meets the need for investigating coupled processes in the subsurface involving regional multi-fluid flow and biogeochemistry. The scalability of eSTOMP makes it ideal for simulating contaminant and geochemical reactive transport when long run times result from execution of a serial code. To demonstrate the impact of high-performance computation on subsurface analyses, the application of eSTOMP is presented for two sites. The first site involves uranium surface complexation within a highly dynamic, variably saturated flow field. The use of this model in conjunction with continued field monitoring is expected to provide a rigorous basis for developing operational strategies for field remediation and for defining defensible remediation endpoints. The second site involves the long-term dissolution of low-level radioactive glass emplaced within a near-surface repository. The waste form release rate is evaluated by modeling the basic physical and chemical processes that are known to control the waste form dissolution behavior, including pH and composition of the fluid contacting the glass. Simulations demonstrate that glass dissolution rates vary both in time and as a function of position within the disposal system, and that a 'leach rate' or radionuclide release rate parameter can only be estimated after identifying these impacts in a dynamic system. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A network approach for multiscale catchment classification using traits

Abstract. The classification of river catchments into groups with similar biophysical characteristics is useful to understand and predict their hydrological behavior. The increasing availability of remote sensing and other large-scale geospatial datasets has enabled the use of advanced data-driven approaches to classify catchments using traits such as topography, geology, climate, land cover, land use, and human influence. Unsupervised clustering algorithms based on the Euclidean distance are commonly used for trait-based classification but are not suitable for highly dimensional data. In this study we present a new network-based method for multi-scale catchment classification, which can be applied to large datasets and used to determine the traits associated with different catchment groups. In this framework, two networks are analyzed in parallel: the first being where the nodes are traits and the second being where the nodes are catchments. In both cases, edges represent pairwise similarity, and a network cluster detection algorithm is used for the classification. The trait network is used to investigate redundancy in the trait data and to condense this information into a small number of interpretable categories. The catchments network is used to classify the catchments into clusters and to identify representative catchments for the different groups using the degree centrality metric. We apply this method to classify 9067 river catchments across the contiguous United States at both regional and continental scales using 274 non-categorical traits. At the continental scale, we identify 25 interpretable trait categories and 34 catchment clusters of sizes greater than 50. We find that catchments with similar trait categories are typically located in the same region, with different spatial patterns emerging among clusters dominated by natural and anthropogenic traits. We also find that the catchment clusters exhibit distinct hydrological behavior based on an analysis of streamflow indices. This network approach provides several advantages over traditional means of classification, including better separation of clusters, the use of alternate similarity metrics that are more suitable for highly dimensional data, and reducing redundancy in the trait information. The paired catchment–trait networks enable analysis of hydrological behavior using the dominant trait categories for each catchment cluster. The approach can be used at multiple spatial scales since the network topologies adjust automatically to reflect the trait patterns at the scale of investigation. Finally, the representative catchments identified as hub nodes in the network can be used to guide transferable observational and modeling strategies. The method is broadly applicable beyond hydrology for classification of other complex systems that utilize different types of trait datasets.

54 ENVIRONMENTAL SCIENCES↗

WHY IS NON-THERMAL LINE BROADENING OF SPECTRAL LINES IN THE LOWER TRANSITION REGION OF THE SUN INDEPENDENT OF SPATIAL RESOLUTION?

Spectral observations of the solar transition region (TR) and corona show broadening of spectral lines beyond what is expected from thermal and instrumental broadening. The remaining non-thermal broadening is significant (5–30 km s{sup −1}) and correlated with intensity. Here we study spectra of the TR Si iv 1403 Å line obtained at high resolution with the Interface Region Imaging Spectrograph (IRIS). We find that the large improvement in spatial resolution (0.″33) of IRIS compared to previous spectrographs (2″) does not resolve the non-thermal line broadening which, in most regions, remains at pre-IRIS levels of about 20 km s{sup −1}. This invariance to spatial resolution indicates that the processes behind the broadening occur along the line-of-sight (LOS) and/or on spatial scales (perpendicular to the LOS) smaller than 250 km. Both effects appear to play a role. Comparison with IRIS chromospheric observations shows that, in regions where the LOS is more parallel to the field, magneto-acoustic shocks driven from below impact the TR and can lead to significant non-thermal line broadening. This scenario is supported by MHD simulations. While these do not show enough non-thermal line broadening, they do reproduce the long-known puzzling correlation between non-thermal line broadening and intensity. This correlation is caused by the shocks, but only if non-equilibrium ionization is taken into account. In regions where the LOS is more perpendicular to the field, the prevalence of small-scale twist is likely to play a significant role in explaining the invariance and correlation with intensity. (letters)

79 ASTRONOMY AND ASTROPHYSICS↗

3D magnetic field configuration of small-scale reconnection events in the solar plasma atmosphere

The outer solar atmosphere, i.e., the corona and the chromosphere, is replete with small energy-release events, which are accompanied by transient brightening and jet-like ejections. These events are considered to be magnetic reconnection events in the solar plasma, and their dynamics have been studied using recent advanced observations from the Hinode spacecraft and other observatories in space and on the ground. These events occur at different locations in the solar atmosphere and vary in their morphology and amount of the released energy. The magnetic field configurations of these reconnection events are inferred based on observations of magnetic fields at the photospheric level. Observations suggest that these magnetic configurations can be classified into two groups. In the first group, two anti-parallel magnetic fields reconnect to each other, yielding a 2D emerging flux configuration. In the second group, helical or twisted magnetic flux tubes are parallel or at a relative angle to each other. Reconnection can occur only between anti-parallel components of the magnetic flux tubes and may be referred to as component reconnection. The latter configuration type may be more important for the larger class of small-scale reconnection events. The two types of magnetic configurations can be compared to counter-helicity and co-helicity configurations, respectively, in laboratory plasma collision experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Distributed structure determination at the JCSG

The software suite Xsolve semi-exhaustively explores key parameters of the X-ray structure-determination process to compute multiple three-dimensional protein structures independently and in parallel from a set of diffraction images. An optimal consensus model for subsequent manual refinement is computed from these structures. The Joint Center for Structural Genomics (JCSG), one of four large-scale structure-determination centers funded by the US Protein Structure Initiative (PSI) through the National Institute for General Medical Sciences, has been operating an automated distributed structure-solution pipeline, Xsolve, for well over half a decade. During PSI-2, Xsolve solved, traced and partially refined 90% of the JCSG’s nearly 770 MAD/SAD structures at an average resolution of about 2 Å without human intervention. Xsolve executes many well established publicly available crystallography software programs in parallel on a commodity Linux cluster, resulting in multiple traces for any given target. Additional software programs have been developed and integrated into Xsolve to further minimize human effort in structure refinement. ConsensusModeler exploits complementarities in traces from Xsolve to compute a single optimal model for manual refinement. Xpleo is a powerful robotics-inspired algorithm to build missing fragments and qFit automatically identifies and fits alternate conformations.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Characteristics of an envelope model for laser-plasma accelerator simulation

Simulation of laser-plasma accelerator (LPA) experiments is computationally intensive due to the disparate length scales involved. Current experiments extend hundreds of laser wavelengths transversely and many thousands in the propagation direction, making explicit PIC simulations enormously expensive and requiring massively parallel execution in 3D. Simulating the next generation of LPA experiments is expected to increase the computational requirements yet further, by a factor of 1000. We can substantially improve the performance of LPA simulations by modeling the envelope evolution of the laser field rather than the field itself. This allows for much coarser grids, since we need only resolve the plasma wavelength and not the laser wavelength, and therefore larger timesteps can be used. Thus an envelope model can result in savings of several orders of magnitude in computational resources. By propagating the laser envelope in a Galilean frame moving at the speed of light, dispersive errors can be avoided and simulations over long distances become possible. The primary limitation to this envelope model is when the laser pulse develops large frequency shifts, and thus the slowly-varying envelope assumption is no longer valid. Here we describe the model and its implementation, and show rigorous benchmarks for the algorithm, establishing second-order convergence and correct laser group velocity. We also demonstrate simulations of LPA phenomena such as self-focusing and meter-scale acceleration stages using the model.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A new scaling for divertor detachment

The ITER design, and future reactor designs, depend on divertor `detachment,'whether partial, pronounced or complete, to limit heat flux to plasma-facing components and to limit surface erosion due to sputtering. It would be valuable to have a measure of the difficulty of achieving detachment as a function of machine parameters, such as input power, magnetic field, major radius, etc. Frequently the parallel heat flux, estimated typically as proportional to P-sep/R or PsepB/R, is used as a proxy for this difficulty. Here we argue that impurity cooling is dependent on the upstream density, which itself must be limited by a Greenwald-like scaling. Taking this into account self-consistently, we find the impurity fraction required for detachment scales dominantly as power divided by poloidal magnetic field. The absence of any explicit scaling with machine size is concerning, as P-sep surely must increase greatly for an economic fusion system, while increases in the poloidal field strength are limited by coil technology and plasma physics. This result should be challenged by comparison with 2D divertor codes and with measurements on existing experiments. Nonetheless, it suggests that higher magnetic field, stronger shaping, double-null operation, `advanced' divertor configurations, as well as alternate means to handle heat flux such as metallic liquid and/or vapor targets merit greater attention.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Refine your search to explore more results.