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 1,459 records · Page 81

Building more powerful less expensive supercomputers using Processing-In-Memory (PIM) LDRD final report.

This report details the accomplishments of the 'Building More Powerful Less Expensive Supercomputers Using Processing-In-Memory (PIM)' LDRD ('PIM LDRD', number 105809) for FY07-FY09. Latency dominates all levels of supercomputer design. Within a node, increasing memory latency, relative to processor cycle time, limits CPU performance. Between nodes, the same increase in relative latency impacts scalability. Processing-In-Memory (PIM) is an architecture that directly addresses this problem using enhanced chip fabrication technology and machine organization. PIMs combine high-speed logic and dense, low-latency, high-bandwidth DRAM, and lightweight threads that tolerate latency by performing useful work during memory transactions. This work examines the potential of PIM-based architectures to support mission critical Sandia applications and an emerging class of more data intensive informatics applications. This work has resulted in a stronger architecture/implementation collaboration between 1400 and 1700. Additionally, key technology components have impacted vendor roadmaps, and we are in the process of pursuing these new collaborations. This work has the potential to impact future supercomputer design and construction, reducing power and increasing performance. This final report is organized as follow: this summary chapter discusses the impact of the project (Section 1), provides an enumeration of publications and other public discussion of the work (Section 1), and concludes with a discussion of future work and impact from the project (Section 1). The appendix contains reprints of the refereed publications resulting from this work.

97 MATHEMATICS AND COMPUTING↗

A Parallel Graph Environment for Real-World Data Analytics Workflows

Economic competitiveness and national security depend increasingly on the insightful analysis of large data sets. The diversity of real-world data sources and analytic workflows impose challenging hardware and software requirements for parallel graph platforms. The irregular nature of graph methods is not supported well by the deep memory hierarchies of conventional distributed systems, requiring new processor and runtime system designs to tolerate memory and synchronization latencies. Moreover, the efficiency of relational table operations and matrix computations are not attainable when data is stored in common graph data structures. In this paper, we present HAGGLE, a high-performance, scalable data analytics platform. The platform’s hybrid data model supports a variety of distributed, thread-safe data structures, parallel programming constructs, and persistent and streaming data. An abstract runtime layer enables us to map the stack to conventional, distributed computer systems with accelerators. The runtime uses multithreading, active messages, and data aggregation to hide memory and synchronization latencies on large-scale systems.

Graph Analytics↗

Assembly, Test and Launch Operations for a Nuclear-enabled NASA Mission: Considerations that are Specific to Use of a Nuclear Payload

For more than five decades, Radioisotope Power Systems (RPS) have played a critical role in the exploration of space, enabling missions of scientific discovery to destinations across the solar system by providing electrical power to explore remote, challenging and extreme environments. In particular, RPS enable deep space missions where increased heliocentric distances reduce the ability of solar power to adequately meet spacecraft and instruments power requirements. Some previous notable missions that were enabled by RPS include Nimbus III, the Apollo Surface Experiments, the Pioneers 10 and 11, the Viking Mars Landers, Galileo, Ulysses, Cassini, New Horizons and Curiosity. The current operating set of missions that are enabled by RPS are Voyagers 1 and 2, Cassini, New Horizons, and Curiosity. The Multi-Mission Radioisotope Thermal Generator (MMRTG) is the current RPS used for missions. An enhanced version of this generator outfitted with higher efficiency thermoelectrics is under development for potential use in the future. Other previously deployed power systems include the Multi-hundred Watt Radisotope Thermo-electric Generator (MHW RTG) and the General Purpose Heat Source Radisotope Thermo-electric Generator (GPHS RTG). The common thread for all of these power systems is that they are fueled with Pu-238. The use of this unique isotope involves additional planning activities and requires specific actions when the devices are delivered to Kennedy Space Center (KSC), launch site, and incorporated into the assembly, test and launch operations (ATLO) process to ensure mission success, met safety and security challenges. It has been forecasted that the use of a nuclear reactor system is on the horizon. This system could be used for either specifically powering S/C propulsion system or for surface power use once the mission arrived at its destina-tion. Since a nuclear-fission-based system has never been handled or integrated into a spacecraft at KSC, this nuclear reactor system would potentially require further challenges that must be accounted for in the ATLO process. This paper will explain ATLO considerations for the recent RTG-enabled launches that have occurred and planned in the near future (Mars 2020 NASA mission). It will also describe where further challenges that need to be met for the placement of a nuclear reactor (either for propulsion or transportation for surface power use) onto a space mission. Specifically, the following topics will be addressed: • Approach for nuclear safety planning for nuclear material use and its transportation to launch site • Planning for security posture for nuclear materials at launch site • Preparation for transportation of the nuclear power system from the fueling/testing location to the launch site • Preparation of documentation and procedures for nuclear material use at launch site • Planning for coordination between power system and space mission groups • Planning for appropriate staffing and scheduling of operations

07 ISOTOPE AND RADIATION SOURCES↗

A UPC++ Actor Library and Its Evaluation On a Shallow Water Proxy Application

Programmability is one of the key challenges of Exascale Computing. Using the actor model for distributed computations may be one solution. The actor model separates computation from communication while still enabling their overlap. Each actor possesses specified communication endpoints to publish and receive information. Computations are undertaken based on the data available on these channels. We present a library that implements this programming model using UPC++, a PGAS library, and evaluate three different parallelization strategies, one based on rank-sequential execution, one based on multiple threads in a rank, and one based on OpenMP tasks. In an evaluation of our library using shallow water proxy applications, our solution compares favorably against an earlier implementation based on X10, and a BSP-based approach.

Pöppl, Alexander↗

Influence of Nut Design on the Load Distribution over the Turns of a Screw

Screw turns are known to be deformed as a result of the contact pliability of the nut and screw surfaces. The load distribution over the turns is investigated when fine thread is employed, when compression nuts are replaced by tension–compression and tension nuts, when the roughness of the working surfaces is increased, and when the load on the screw is increased.

42 ENGINEERING↗

GPU-Accelerated Machine Learning Inference as a Service for Computing in Neutrino Experiments

Machine learning algorithms are becoming increasingly prevalent and performant in the reconstruction of events in accelerator-based neutrino experiments. These sophisticated algorithms can be computationally expensive. At the same time, the data volumes of such experiments are rapidly increasing. The demand to process billions of neutrino events with many machine learning algorithm inferences creates a computing challenge. We explore a computing model in which heterogeneous computing with GPU coprocessors is made available as a web service. The coprocessors can be efficiently and elastically deployed to provide the right amount of computing for a given processing task. With our approach, Services for Optimized Network Inference on Coprocessors (SONIC), we integrate GPU acceleration specifically for the ProtoDUNE-SP reconstruction chain without disrupting the native computing workflow. With our integrated framework, we accelerate the most time-consuming task, track and particle shower hit identification, by a factor of 17. This results in a factor of 2.7 reduction in the total processing time when compared with CPU-only production. For this particular task, only 1 GPU is required for every 68 CPU threads, providing a cost-effective solution.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

The PENELOPE Physics Models and Transport Mechanics. Implementation into Geant4

A translation of the PENELOPE physics subroutines to C++, designed as an extension of the GEANT4 toolkit, is presented. The Fortran code system PENELOPE performs Monte Carlo simulation of coupled electron-photon transport in arbitrary materials for a wide energy range, nominally from 50 eV up to 1 GeV. PENELOPE implements the most reliable interaction models that are currently available, limited only by the required generality of the code. In addition, the transport of electrons and positrons is simulated by means of an elaborate class II scheme in which hard interactions (involving deflection angles or energy transfers larger than pre-defined cutoffs) are simulated from the associated restricted differential cross sections. After a brief description of the interaction models adopted for photons and electrons/positrons, we describe the details of the class-II algorithm used for tracking electrons and positrons. The C++ classes are adapted to the specific code structure of GEANT4. They provide a complete description of the interactions and transport mechanics of electrons/positrons and photons in arbitrary materials, which can be activated from the G4ProcessManager to produce simulation results equivalent to those from the original PENELOPE programs. The combined code, named PENG4, benefits from the multi-threading capabilities and advanced geometry and statistical tools of GEANT4.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Machine Learning-Driven Conservative-to-Primitive Conversion in Hybrid Piecewise Polytropic and Tabulated Equations of State

We present a novel machine learning (ML)-based method to accelerate conservative-to-primitive inversion, focusing on hybrid piecewise polytropic and tabulated equations of state. Traditional root-finding techniques are computationally expensive, particularly for large-scale relativistic hydrodynamics simulations. To address this, we employ feedforward neural networks (NNC2PS and NNC2PL), trained in PyTorch (2.0+) and optimized for GPU inference using NVIDIA TensorRT (8.4.1), achieving significant speedups with minimal accuracy loss. The NNC2PS model achieves 𝐿 1 and 𝐿 ∞ errors of 4.54 × 10 −7 and 3.44 × 10−6, respectively, while the NNC2PL model exhibits even lower error values. TensorRT optimization with mixed-precision deployment substantially accelerates performance compared to traditional root-finding methods. Specifically, the mixed-precision TensorRT engine for NNC2PS achieves inference speeds approximately 400 times faster than a traditional single-threaded CPU implementation for a dataset size of 1,000,000 points. Ideal parallelization across an entire compute node in the Delta supercomputer (dual AMD 64-core 2.45 GHz Milan processors and 8 NVIDIA A100 GPUs with 40 GB HBM2 RAM and NVLink) predicts a 25-fold speedup for TensorRT over an optimally parallelized numerical method when processing 8 million data points. Moreover, the ML method exhibits sub-linear scaling with increasing dataset sizes. We release the scientific software developed, enabling further validation and extension of our findings. By exploiting the underlying symmetries within the equation of state, these findings highlight the potential of ML, combined with GPU optimization and model quantization, to accelerate conservative-to-primitive inversion in relativistic hydrodynamics simulations.

conservative-to-primitive conversion↗

High figure of merit extreme bandgap Al 0.87 Ga 0.13 N-Al 0.64 Ga 0.36 N heterostructures over bulk AlN substrates

We report on high-quality n -Al 0.87 Ga 0.13 N-A 0.64 Ga 0.36 N heterostructures over single crystal AlN. For these pseudomorphic heterostructures, high-resolution X-ray and X-ray Topographic analysis was used to establish a threading dislocation density of 7 × 10 3 cm -2 . Using reverse composition graded n + -Al x Ga 1- x N contact layers, we obtained linear ohmic contacts with 4.3 Ω mm specific resistance. A critical breakdown field >11 MV cm -1 was also measured. In combination with the channel resistance of 2400 Ω sq -1 , these translate to a Baliga’s Figure of Merit of 2.27 GW cm -2 . This, to the best of our knowledge is the highest reported value for extreme bandgap AlGaN heterostructures.

42 ENGINEERING↗

High-Temperature and High-Pressure In Situ Magic Angle Spinning Nuclear Magnetic Resonance Spectroscopy

Nuclear magnetic resonance (NMR) spectroscopy represents an important technique to understand the structure and bonding environment of molecules. There exists a drive to conduct characterization experiment under conditions relevant to the chemical process of interest. To address this, in situ high-temperature, high-pressure MAS NMR methods have been developed to enable the observation of chemical interactions over a range of pressures (vacuum to several hundred bar) and temperatures (well below 0°C to 250°C). Further, the chemical identity of the samples can be comprised of solids, liquids, and gases or mixtures of the three. The method incorporates all-zirconia NMR rotors (sample holder for MAS NMR) which can be sealed using a threaded cap to compress an O-ring. This rotor exhibits great chemical resistance, temperature compatibility, low NMR background, and can withstand high pressures. These factors combined enable its use in a wide range of system combinations which permit its use in diverse fields as carbon sequestration, catalysis, material science, geochemistry, and biology. The flexibility of this technique makes it an attractive option for scientists from numerous disciplines.

Nuclear Magnetic Resonance, NMR, in situ, spectros↗

Vertical structure of magnetized accretion disks around young stars

We model the vertical structure of the magnetized accretion disks that are subject to viscous and resistive heating and irradiation by the central star. We apply our formalism to the radial structure of the magnetized accretion disks that are threaded by the poloidal magnetic field dragged during the process of star formation, which was developed by Shu and coworkers. We consider disks around low-mass protostars, T Tauri, and FU Orionis stars, as well as two levels of disk magnetization: λ{sub sys}=4 (strongly magnetized disks) and λ{sub sys}=12 (weakly magnetized disks). The rotation rates of strongly magnetized disks have large deviations from Keplerian rotation. In these models, resistive heating dominates the thermal structure for the FU Ori disk, and the T Tauri disk is very thin and cold because it is strongly compressed by magnetic pressure; it may be too thin compared with observations. Instead, in the weakly magnetized disks, rotation velocities are close to Keplerian, and resistive heating is always less than 7% of the viscous heating. In these models, the T Tauri disk has a larger aspect ratio, which is consistent with that inferred from observations. All the disks have spatially extended hot atmospheres where the irradiation flux is absorbed, although most of the mass (∼90%–95%) is in the disk midplane. With the advent of ALMA one expects direct measurements of magnetic fields and their morphology at disk scales. It will then be possible to determine the mass-to-flux ratio of magnetized accretion disks around young stars, an essential parameter for their structure and evolution. Our models contribute to the understanding of the vertical structure and emission of these disks.

79 ASTRONOMY AND ASTROPHYSICS↗

An accretion disk-outflow model for hysteretic state transition in X-ray binaries

We suggest a model of the advection-dominated accretion flow (ADAF) with magnetically driven outflows to explain the hysteretic state transition observed in X-ray binaries (XRBs). The transition from a thin disk to an ADAF occurs when the mass accretion rate is below a critical value. The critical mass accretion rate for the ADAF can be estimated by equating the equilibration timescale to the accretion timescale of the ADAF, which is sensitive to its radial velocity. The radial velocity of thin disks is very small, which leads to the advection of the external field in thin disks becoming very inefficient. ADAFs are present in the low/hard states of XRBs, and their radial velocity is large compared with the thin disk. The external field can be dragged inward efficiently by the ADAF, so a strong large-scale magnetic field threading the ADAF can be formed, which may accelerate a fraction of gas in the ADAF into the outflows. Such outflows may carry away a large amount of angular momentum from the ADAF, which significantly increases the radial velocity of the ADAF. This leads to a high critical mass accretion rate, below which an ADAF with magnetic outflows can survive. Our calculations show that the critical luminosity of the ADAF with magnetic outflows can be one order of magnitude higher than that for a conventional ADAF, if the ratio of gas to magnetic pressure β∼4 in the disk. This can naturally explain the hysteretic state transition observed in XRBs.

79 ASTRONOMY AND ASTROPHYSICS↗

Probing the innermost regions of AGN jets and their magnetic fields with RadioAstron. I. Imaging BL Lacertae at 21 μas resolution

We present the first polarimetric space very long baseline interferometry (VLBI) imaging observations at 22 GHz. BL Lacertae was observed in 2013 November 10 with the RadioAstron space VLBI mission, including a ground array of 15 radio telescopes. The instrumental polarization of the space radio telescope is found to be less than 9%, demonstrating the polarimetric imaging capabilities of RadioAstron at 22 GHz. Ground–space fringes were obtained up to a projected baseline distance of 7.9 Earth diameters in length, allowing us to image the jet in BL Lacertae with a maximum angular resolution of 21 μas, the highest achieved to date. We find evidence for emission upstream of the radio core, which may correspond to a recollimation shock at about 40 μas from the jet apex, in a pattern that includes other recollimation shocks at approximately 100 and 250 μas from the jet apex. Polarized emission is detected in two components within the innermost 0.5 mas from the core, as well as in some knots 3 mas downstream. Faraday rotation analysis, obtained from combining RadioAstron 22 GHz and ground-based 15 and 43 GHz images, shows a gradient in rotation measure and Faraday-corrected polarization vector as a function of position angle with respect to the core, suggesting that the jet in BL Lacertae is threaded by a helical magnetic field. The intrinsic de-boosted brightness temperature in the unresolved core exceeds 3×10{sup 12} K, suggesting, at the very least, departure from equipartition of energy between the magnetic field and radiating particles.

79 ASTRONOMY AND ASTROPHYSICS↗

ATMOSPHERIC CIRCULATION OF HOT JUPITERS: DAYSIDE–NIGHTSIDE TEMPERATURE DIFFERENCES

The full-phase infrared light curves of low-eccentricity hot Jupiters show a trend of increasing dayside-to-nightside brightness temperature difference with increasing equilibrium temperature. Here, we present a three-dimensional model that explains this relationship, in order to provide insight into the processes that control heat redistribution in tidally locked planetary atmospheres. This three-dimensional model combines predictive analytic theory for the atmospheric circulation and dayside–nightside temperature differences over a range of equilibrium temperatures, atmospheric compositions, and potential frictional drag strengths with numerical solutions of the circulation that verify this analytic theory. The theory shows that the longitudinal propagation of waves mediates dayside–nightside temperature differences in hot Jupiter atmospheres, analogous to the wave adjustment mechanism that regulates the thermal structure in Earth’s tropics. These waves can be damped in hot Jupiter atmospheres by either radiative cooling or potential frictional drag. This frictional drag would likely be caused by Lorentz forces in a partially ionized atmosphere threaded by a background magnetic field, and would increase in strength with increasing temperature. Additionally, the amplitude of radiative heating and cooling increases with increasing temperature, and hence both radiative heating/cooling and frictional drag damp waves more efficiently with increasing equilibrium temperature. Radiative heating and cooling play the largest role in controlling dayside–nightside temperature differences in both our analytic theory and numerical simulations, with frictional drag only being important if it is stronger than the Coriolis force. As a result, dayside–nightside temperature differences in hot Jupiter atmospheres increase with increasing stellar irradiation and decrease with increasing pressure.

79 ASTRONOMY AND ASTROPHYSICS↗

TOWARD A GLOBAL EVOLUTIONARY MODEL OF PROTOPLANETARY DISKS

A global picture of the evolution of protoplanetary disks (PPDs) is key to understanding almost every aspect of planet formation, where standard α-disk models have been continually employed for their simplicity. In the meantime, disk mass loss has been conventionally attributed to photoevaporation, which controls disk dispersal. However, a paradigm shift toward accretion driven by magnetized disk winds has taken place in recent years, thanks to studies of non-ideal magnetohydrodynamic effects in PPDs. I present a framework of global PPD evolution aiming to incorporate these advances, highlighting the role of wind-driven accretion and wind mass loss. Disk evolution is found to be largely dominated by wind-driven processes, and viscous spreading is suppressed. The timescale of disk evolution is controlled primarily by the amount of external magnetic flux threading the disks, and how rapidly the disk loses the flux. Rapid disk dispersal can be achieved if the disk is able to hold most of its magnetic flux during the evolution. In addition, because wind launching requires a sufficient level of ionization at the disk surface (mainly via external far-UV (FUV) radiation), wind kinematics is also affected by the FUV penetration depth and disk geometry. For a typical disk lifetime of a few million years, the disk loses approximately the same amount of mass through the wind as through accretion onto the protostar, and most of the wind mass loss proceeds from the outer disk via a slow wind. Fractional wind mass loss increases with increasing disk lifetime. Significant wind mass loss likely substantially enhances the dust-to-gas mass ratio and promotes planet formation.

79 ASTRONOMY AND ASTROPHYSICS↗

THREE-DIMENSIONAL SIMULATIONS OF TEARING AND INTERMITTENCY IN CORONAL JETS

Observations of coronal jets increasingly suggest that local fragmentation and intermittency play an important role in the dynamics of these events. In this work, we investigate this fragmentation in high-resolution simulations of jets in the closed-field corona. We study two realizations of the embedded-bipole model, whereby impulsive helical outflows are driven by reconnection between twisted and untwisted field across the domed fan plane of a magnetic null. We find that the reconnection region fragments following the onset of a tearing-like instability, producing multiple magnetic null points and flux-rope structures within the current layer. The flux ropes formed within the weak-field region in the center of the current layer are associated with “blobs” of density enhancement that become filamentary threads as the flux ropes are ejected from the layer, whereupon new flux ropes form behind them. This repeated formation and ejection of flux ropes provides a natural explanation for the intermittent outflows, bright blobs of emission, and filamentary structure observed in some jets. Additional observational signatures of this process are discussed. Essentially all jet models invoke reconnection between regions of locally closed and locally open field as the jet-generation mechanism. Therefore, we suggest that this repeated tearing process should occur at the separatrix surface between the two flux systems in all jets. A schematic picture of tearing-mediated jet reconnection in three dimensions is outlined.

79 ASTRONOMY AND ASTROPHYSICS↗

FLUX CANCELLATION AND THE EVOLUTION OF THE ERUPTIVE FILAMENT OF 2011 JUNE 7

We investigate whether flux cancellation is responsible for the formation of a very massive filament resulting in the spectacular eruption on 2011 June 7. We analyze and quantify the amount of flux cancellation that occurs in NOAA AR 11226 and its two neighboring active regions (ARs 11227 and 11233) using line-of-sight magnetograms from the Heliospheric Magnetic Imager. During a 3.6 day period building up to the eruption of the filament, 1.7 × 10{sup 21} Mx, 21% of AR 11226's maximum magnetic flux, was canceled along the polarity inversion line (PIL) where the filament formed. If the flux cancellation continued at the same rate up until the eruption then up to 2.8 × 10{sup 21} Mx (34% of the AR flux) may have been built into the magnetic configuration that contains the filament plasma. The large flux cancellation rate is due to an unusual motion of the positive-polarity sunspot, which splits, with the largest section moving rapidly toward the PIL. This motion compresses the negative polarity and leads to the formation of an orphan penumbra where one end of the filament is rooted. Dense plasma threads above the orphan penumbra build into the filament, extending its length, and presumably injecting material into it. We conclude that the exceptionally strong flux cancellation in AR 11226 played a significant role in the formation of its unusually massive filament. In addition, the presence and coherent evolution of bald patches in the vector magnetic field along the PIL suggest that the magnetic field configuration supporting the filament material is that of a flux rope.

79 ASTRONOMY AND ASTROPHYSICS↗

The magnetic furnace: intense core dynamos in B stars

The dynamo action achieved in the convective cores of main-sequence massive stars is explored here through three-dimensional (3D) global simulations of convective core dynamos operating within a young 10 M{sub ⊙} B-type star, using the anelastic spherical harmonic code. These simulations capture the inner 65% of this star by radius, encompassing the convective nuclear-burning core (about 23% by radius) and a portion of the overlying radiative envelope. Eight rotation rates are considered, ranging from 0.05% to 16% of the surface breakup velocity, thereby capturing both convection that barely senses the effects of rotation and other situations in which the Coriolis forces are prominent. The vigorous dynamo action realized within all of these turbulent convective cores builds magnetic fields with peak strengths exceeding a megagauss, with the overall magnetic energy (ME) in the faster rotators reaching super-equipartition levels compared to the convective kinetic energy (KE). The core convection typically involves turbulent columnar velocity structures roughly aligned with the rotation axis, with magnetic fields threading through these rolls and possessing complex linkages throughout the core. The very strong fields are able to coexist with the flows without quenching them through Lorentz forces. The velocity and magnetic fields achieve such a state by being nearly co-aligned, and with peak magnetic islands being somewhat displaced from the fastest flows as the intricate evolution proceeds. As the rotation rate is increased, the primary force balance shifts from nonlinear advection balancing Lorentz forces to a magnetostrophic balance between Coriolis and Lorentz forces.

79 ASTRONOMY AND ASTROPHYSICS↗