Search NASA⌕ Search

SEARCH · Search NASA

Results for “Threading”

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 649 records · Page 36

UPC++ v1.0 Specification, Revision 2023.3.0

UPC++ is a C++ library providing classes and functions that support Partitioned Global Address Space (PGAS) programming. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). All communication operations are syntactically explicit and default to non-blocking; asynchrony is managed through the use of futures, promises and continuation callbacks, enabling the programmer to construct a graph of operations to execute asynchronously as high-latency dependencies are satisfied. A global pointer abstraction provides system-wide addressability of shared memory, including host and accelerator memories. The parallelism model is primarily process-based, but the interface is thread-safe and designed to allow efficient and expressive use in multi-threaded applications. The interface is designed for extreme scalability throughout, and deliberately avoids design features that could inhibit scalability.

97 MATHEMATICS AND COMPUTING↗

UPC++ v1.0 Specification (Revision 2021.3.0)

UPC++ is a C++11 library providing classes and functions that support Partitioned Global Address Space (PGAS) programming. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). All communication operations are syntactically explicit and default to non-blocking; asynchrony is managed through the use of futures, promises and continuation callbacks, enabling the programmer to construct a graph of operations to execute asynchronously as high-latency dependencies are satisfied. A global pointer abstraction provides system-wide addressability of shared memory, including host and accelerator memories. The parallelism model is primarily process-based, but the interface is thread-safe and designed to allow efficient and expressive use in multi-threaded applications. The interface is designed for extreme scalability throughout, and deliberately avoids design features that could inhibit scalability.

97 MATHEMATICS AND COMPUTING↗

UPC++ v1.0 Specification, Revision 2022.9.0

UPC++ is a C++ library providing classes and functions that support Partitioned Global Address Space (PGAS) programming. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). All communication operations are syntactically explicit and default to non-blocking; asynchrony is managed through the use of futures, promises and continuation callbacks, enabling the programmer to construct a graph of operations to execute asynchronously as high-latency dependencies are satisfied. A global pointer abstraction provides system-wide addressability of shared memory, including host and accelerator memories. The parallelism model is primarily process-based, but the interface is thread-safe and designed to allow efficient and expressive use in multi-threaded applications. The interface is designed for extreme scalability throughout, and deliberately avoids design features that could inhibit scalability.

97 MATHEMATICS AND COMPUTING↗

UPC++ v1.0 Specification, Revision 2020.3.0

UPC++ is a C++11 library providing classes and functions that support Partitioned Global Address Space (PGAS) programming. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). All communication operations are syntactically explicit and default to non-blocking; asynchrony is managed through the use of futures, promises and continuation callbacks, enabling the programmer to construct a graph of operations to execute asynchronously as high-latency dependencies are satisfied. A global pointer abstraction provides system-wide addressability of shared memory, including host and accelerator memories. The parallelism model is primarily process-based, but the interface is thread-safe and designed to allow efficient and expressive use in multi-threaded applications. The interface is designed for extreme scalability throughout, and deliberately avoids design features that could inhibit scalability.

97 MATHEMATICS AND COMPUTING↗

OBSERVATIONS OF A SERIES OF FLARES AND ASSOCIATED JET-LIKE ERUPTIONS DRIVEN BY THE EMERGENCE OF TWISTED MAGNETIC FIELDS

We studied temporal changes of morphological and magnetic properties of a succession of four confined flares followed by an eruptive flare using the high-resolution New Solar Telescope (NST) operating at the Big Bear Solar Observatory (BBSO) and Helioseismic and Magnetic Imager (HMI) magnetograms and Atmospheric Image Assembly (AIA) EUV images provided by the Solar Dynamics Observatory (SDO). From the NST/Hα and the SDO/AIA 304 Å observations we found that each flare developed a jet structure that evolved in a manner similar to evolution of the blowout jet: (1) an inverted-Y-shaped jet appeared and drifted away from its initial position; (2) jets formed a curtain-like structure that consisted of many fine threads accompanied by subsequent brightenings near the footpoints of the fine threads; and finally, (3) the jet showed a twisted structure visible near the flare maximum. Analysis of the HMI data showed that both the negative magnetic flux and the magnetic helicity have been gradually increasing in the positive-polarity region, indicating the continuous injection of magnetic twist before and during the series of flares. Based on these results, we suggest that the continuous emergence of twisted magnetic flux played an important role in producing successive flares and developing a series of blowout jets.

79 ASTRONOMY AND ASTROPHYSICS↗

FORMATION AND PLASMA CIRCULATION OF SOLAR PROMINENCES

Solar prominences are long-lived cool and dense plasma curtains in the hot and rarefied outer solar atmosphere or corona. The physical mechanism responsible for their formation and especially for their internal plasma circulation has been uncertain for decades. The observed ubiquitous downflows in quiescent prominences are difficult to interpret because plasma with high conductivity seems to move across horizontal magnetic field lines. Here we present three-dimensional numerical simulations of prominence formation and evolution in an elongated magnetic flux rope as a result of in situ plasma condensations fueled by continuous plasma evaporation from the solar chromosphere. The prominence is born and maintained in a fragmented, highly dynamic state with continuous reappearance of multiple blobs and thread structures that move mainly downward, dragging along mass-loaded field lines. The circulation of prominence plasma is characterized by the dynamic balance between the drainage of prominence plasma back to the chromosphere and the formation of prominence plasma via continuous condensation. Plasma evaporates from the chromosphere, condenses into the prominence in the corona, and drains back to the chromosphere, establishing a stable chromosphere–corona plasma cycle. Synthetic images of the modeled prominence with the Solar Dynamics Observatory Atmospheric Imaging Assembly closely resemble actual observations, with many dynamical threads underlying an elliptical coronal cavity.

79 ASTRONOMY AND ASTROPHYSICS↗

THE INTERNAL STRUCTURE OF OVERPRESSURED, MAGNETIZED, RELATIVISTIC JETS

This work presents the first characterization of the internal structure of overpressured, steady superfast-magnetosonic relativistic jets in connection with their dominant type of energy. To this aim, relativistic magnetohydrodynamic simulations of different jet models threaded by a helical magnetic field have been analyzed covering a wide region in the magnetosonic Mach number–specific internal energy plane. The merit of this plane is that models dominated by different types of energy (internal energy: hot jets; rest-mass energy: kinetically dominated jets; magnetic energy: Poynting-flux-dominated jets) occupy well-separated regions. The analyzed models also cover a wide range of magnetizations. Models dominated by the internal energy (i.e., hot models, or Poynting-flux-dominated jets with magnetizations larger than but close to one) have a rich internal structure characterized by a series of recollimation shocks and present the largest variations in the flow Lorentz factor (and internal energy density). Conversely, in kinetically dominated models, there is not much internal or magnetic energy to be converted into kinetic, and the jets are featureless with small variations in the flow Lorentz factor. The presence of a significant toroidal magnetic field threading the jet produces large gradients in the transversal profile of the internal energy density. Poynting-flux-dominated models with high magnetization (≈10 or larger) are prone to be unstable against magnetic pinch modes, which sets limits on the expected magnetization in parsec-scale active galactic nucleus jets or constrains their magnetic field configuration.

79 ASTRONOMY AND ASTROPHYSICS↗

Large-amplitude Longitudinal Oscillations in a Solar Filament

In this paper, we report our multiwavelength observations of the large-amplitude longitudinal oscillations of a filament observed on 2015 May 3. Located next to active region 12335, the sigmoidal filament was observed by the ground-based H α telescopes from the Global Oscillation Network Group and by the Atmospheric Imaging Assembly instrument on board the Solar Dynamics Observatory . The filament oscillations were most probably triggered by the magnetic reconnection in the filament channel, which is characterized by the bidirectional flows, brightenings in EUV and soft X-ray, and magnetic cancellation in the photosphere. The directions of oscillations have angles of 4°–36° with respect to the filament axis. The whole filament did not oscillate in phase as a rigid body. Meanwhile, the oscillation periods (3100–4400 s) have a spatial dependence, implying that the curvature radii ( R ) of the magnetic dips are different at different positions. The values of R are estimated to be 69.4–133.9 Mm, and the minimum transverse magnetic field of the dips is estimated to be 15 G. The amplitudes of S5-S8 grew with time, while the amplitudes of S9-S14 damped with time. The oscillation amplitudes range from a few to ten Mm, and the maximum velocity can reach 30 km s{sup −1}. Interestingly, the filament experienced mass drainage southward at a speed of ∼27 km s{sup −1}. The oscillations continued after the mass drainage and lasted for more than 11 hr. After the mass drainage, the oscillation phases did not change much. The periods of S5-S8 decreased, while the periods of S9-S14 increased. The amplitudes of S5-S8 damped with time, while the amplitudes of S9-S14 grew. Most of the damping (growing) ratios are between −9 and 14. We offer a schematic cartoon to explain the complex behaviors of oscillations by introducing thread-thread interaction.

79 ASTRONOMY AND ASTROPHYSICS↗

WOMBAT: A Scalable and High-performance Astrophysical Magnetohydrodynamics Code

We present a new code for astrophysical magnetohydrodynamics specifically designed and optimized for high performance and scaling on modern and future supercomputers. We describe a novel hybrid OpenMP/MPI programming model that emerged from a collaboration between Cray, Inc. and the University of Minnesota. This design utilizes MPI-RMA optimized for thread scaling, which allows the code to run extremely efficiently at very high thread counts ideal for the latest generation of multi-core and many-core architectures. Such performance characteristics are needed in the era of “exascale” computing. We describe and demonstrate our high-performance design in detail with the intent that it may be used as a model for other, future astrophysical codes intended for applications demanding exceptional performance.

79 ASTRONOMY AND ASTROPHYSICS↗

Aspects of Dislocation Behavior in SiC

A review is presented of the current understanding of the dislocation configurations observed in PVT-grown 4H- and 6H-SiC boules and CVD-grown 4H-SiC homoepitaxial layers. In both PVT-grown boules and CVD-grown epilayers, dislocation configurations are classified according to whether they are growth dislocations, i.e., formed during growth via the replication of dislocations which thread the moving crystal growth front, or result from deformation processes (under either mechanical or electrical stress) immediately following growth, during post growth cooling, i.e., behind the crystal growth front or during device operation. Possible formation mechanisms of growth defects in the PVT grown boules, such as axial screw dislocations and threading edge dislocation walls are proposed. Similarly, possible origins of growth defect configurations in CVD-grown epilayers, such as Frank faults bounded by Frank partials, BPDs and TEDs, are also discussed. In a similar way, the origins of BPD configurations resulting from relaxation of thermal stresses during post-growth cooling of the PVT boules are discussed. Finally, the susceptibility of BPD configurations replicated into CVD grown epilayers from the substrate towards Recombination Enhanced Dislocation Glide (REDG) is discussed.

36 MATERIALS SCIENCE↗

Nucleation of c-axis Screw Dislocations at Substrate Surface Damage During 4H-silicon Carbide Homo-epitaxy

Observations of dislocation nucleation occurring at substrate surface scratches during 4H-SiC CVD homoepitaxial growth are reported. Sub-surface residual damage associated with the scratches is observed to act as nucleation sites for basal plane dislocations (BPDs), threading edge dislocations (TEDs) and threading screw dislocations (TSDs) in the epilayer. TEDs and BPDs replicate from the surface intersections of basal plane dislocation half-loops injected into the substrate surface. A model for the nucleation mechanism of TSDs, which nucleate in opposite sign pairs, is presented which involves overgrowth of surface indentations associated with the scratch during step flow growth. Atomic steps which approach these local surface indentations can collapse creating pairs of opposite sign screw dislocations which have Burgers vector magnitude equal to the magnitude of the step disregistry created during the collapse.

36 MATERIALS SCIENCE↗

Using OpenMP for HEP Framework Algorithm Scheduling [Slides]

The OpenMP standard is the primary mechanism used at high performance computing facilities to allow intra-process parallelization. In contrast, many HEP specific software (such as CMSSW, GaudiHive, and ROOT) make use of Intel's Threading Building Blocks (TBB) library to accomplish the same goal. In this talk we will discuss our work to compare TBB and OpenMP when used for scheduling algorithms to be run by a HEP style data processing framework (i.e. running hundreds of interdependent algorithms at most once for each event read from the detector). This includes both scheduling of different algorithms to be run concurrently as well as scheduling concurrent work within one algorithm. As part of the discussion we present an overview of the OpenMP threading model. We also explain how we used OpenMP when creating a simplified HEP-like processing framework. Using that simplified framework, and a similar one written using TBB, we will present performance comparisons between TBB and different compiler versions of OpenMP.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Fault Diagnosis of Hybrid Computing Systems Using Chaotic-Map Method

Computing systems are becoming increasingly complex with nodes consisting of a combination of multi-core central processing units (CPUs), many integrated core (MIC) and graphics processing unit (GPU) accelerators. These computing units and their interconnections are subject to different classes of hardware and software faults, which should be detected to support mitigation measures. We present the chaotic-map method that uses the exponential divergence and wide Fourier properties of the trajectories, combined with memory allocations and assignments to diagnose component-level faults in these hybrid computing systems. We propose lightweight codes that utilize highly parallel chaotic-map computations tailored to isolate faults in arithmetic units, memory elements and interconnects. The diagnosis module on a node utilizes pthreads to place chaotic-map threads on CPU and MIC cores, and CUDA C and OpenCL kernels on GPU blocks. We present experimental diagnosis results on five multi-core CPUs; one MIC; and, seven GPUs with typical diagnosis run-times under a minute.

Rao, Nageswara↗

Dna2 nuclease-helicase structure, mechanism and regulation by Rpa

The Dna2 nuclease-helicase maintains genomic integrity by processing DNA double-strand breaks, Okazaki fragments and stalled replication forks. Dna2 requires ssDNA ends, and is dependent on the ssDNA-binding protein Rpa, which controls cleavage polarity. Here we present the 2.3 Å structure of intact mouse Dna2 bound to a 15-nucleotide ssDNA. The nuclease active site is embedded in a long, narrow tunnel through which the DNA has to thread. The helicase domain is required for DNA binding but not threading. We also present the structure of a flexibly-tethered Dna2-Rpa interaction that recruits Dna2 to Rpa-coated DNA. We establish that a second Dna2-Rpa interaction is mutually exclusive with Rpa-DNA interactions and mediates the displacement of Rpa from ssDNA. This interaction occurs at the nuclease tunnel entrance and the 5’ end of the Rpa-DNA complex. Hence, it only displaces Rpa from the 5’ but not 3’ end, explaining how Rpa regulates cleavage polarity.

59 BASIC BIOLOGICAL SCIENCES↗

Factors impacting performance of multithreaded sparse triangular solve.

As computational science applications grow more parallel with multi-core supercomputers having hundreds of thousands of computational cores, it will become increasingly difficult for solvers to scale. Our approach is to use hybrid MPI/threaded numerical algorithms to solve these systems in order to reduce the number of MPI tasks and increase the parallel efficiency of the algorithm. However, we need efficient threaded numerical kernels to run on the multi-core nodes in order to achieve good parallel efficiency. In this paper, we focus on improving the performance of a multithreaded triangular solver, an important kernel for preconditioning. We analyze three factors that affect the parallel performance of this threaded kernel and obtain good scalability on the multi-core nodes for a range of matrix sizes.

99 GENERAL AND MISCELLANEOUS↗

Factors impacting performance of multithreaded triangular solve.

As computational science applications grow more parallel with multi-core supercomputers having hundreds of thousands of computational cores, it will become increasingly difficult for solvers to scale. Our approach is to use hybrid MPI/threaded numerical algorithms to solve these systems in order to reduce the number of MPI tasks and increase the parallel efficiency of the algorithm. However, we need efficient threaded numerical kernels to run on the multi-core nodes in order to achieve good parallel efficiency. In this paper, we focus on improving the performance of a multithreaded triangular solver, an important kernel for preconditioning. We analyze three factors that affect the parallel performance of this threaded kernel and obtain good scalability on the multi-core nodes for a range of matrix sizes.

99 GENERAL AND MISCELLANEOUS↗

Dynamic tensile characterization of a 4330-V steel with kolsky bar techniques.

There has been increasing demand to understand the stress-strain response as well as damage and failure mechanisms of materials under impact loading condition. Dynamic tensile characterization has been an efficient approach to acquire satisfactory information of mechanical properties including damage and failure of the materials under investigation. However, in order to obtain valid experimental data, reliable tensile experimental techniques at high strain rates are required. This includes not only precise experimental apparatus but also reliable experimental procedures and comprehensive data interpretation. Kolsky bar, originally developed by Kolsky in 1949 [1] for high-rate compressive characterization of materials, has been extended for dynamic tensile testing since 1960 [2]. In comparison to Kolsky compression bar, the experimental design of Kolsky tension bar has been much more diversified, particularly in producing high speed tensile pulses in the bars. Moreover, instead of directly sandwiching the cylindrical specimen between the bars in Kolsky bar compression bar experiments, the specimen must be firmly attached to the bar ends in Kolsky tensile bar experiments. A common method is to thread a dumbbell specimen into the ends of the incident and transmission bars. The relatively complicated striking and specimen gripping systems in Kolsky tension bar techniques often lead to disturbance in stress wave propagation in the bars, requiring appropriate interpretation of experimental data. In this study, we employed a modified Kolsky tension bar, newly developed at Sandia National Laboratories, Livermore, CA, to explore the dynamic tensile response of a 4330-V steel. The design of the new Kolsky tension bar has been presented at 2010 SEM Annual Conference [3]. Figures 1 and 2 show the actual photograph and schematic of the Kolsky tension bar, respectively. As shown in Fig. 2, the gun barrel is directly connected to the incident bar with a coupler. The cylindrical striker set inside the gun barrel is launched to impact on the end cap that is threaded into the open end of the gun barrel, producing a tension on the gun barrel and the incident bar.

36 MATERIALS SCIENCE↗

Dynamic tensile characterization of a 4330 steel with kolsky bar techniques.

There has been increasing demand to understand the stress-strain response as well as damage and failure mechanisms of materials under impact loading condition. Dynamic tensile characterization has been an efficient approach to acquire satisfactory information of mechanical properties including damage and failure of the materials under investigation. However, in order to obtain valid experimental data, reliable tensile experimental techniques at high strain rates are required. This includes not only precise experimental apparatus but also reliable experimental procedures and comprehensive data interpretation. Kolsky bar, originally developed by Kolsky in 1949 [1] for high-rate compressive characterization of materials, has been extended for dynamic tensile testing since 1960 [2]. In comparison to Kolsky compression bar, the experimental design of Kolsky tension bar has been much more diversified, particularly in producing high speed tensile pulses in the bars. Moreover, instead of directly sandwiching the cylindrical specimen between the bars in Kolsky bar compression bar experiments, the specimen must be firmly attached to the bar ends in Kolsky tensile bar experiments. A common method is to thread a dumbbell specimen into the ends of the incident and transmission bars. The relatively complicated striking and specimen gripping systems in Kolsky tension bar techniques often lead to disturbance in stress wave propagation in the bars, requiring appropriate interpretation of experimental data. In this study, we employed a modified Kolsky tension bar, newly developed at Sandia National Laboratories, Livermore, CA, to explore the dynamic tensile response of a 4330-V steel. The design of the new Kolsky tension bar has been presented at 2010 SEM Annual Conference [3]. Figures 1 and 2 show the actual photograph and schematic of the Kolsky tension bar, respectively. As shown in Fig. 2, the gun barrel is directly connected to the incident bar with a coupler. The cylindrical striker set inside the gun barrel is launched to impact on the end cap that is threaded into the open end of the gun barrel, producing a tension on the gun barrel and the incident bar.

36 MATERIALS SCIENCE↗