Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallelization efficiency”

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 1,675 records · Page 93

In-depth analysis on parallel processing patterns for high-performance Dataframes

The Data Science domain has expanded monumentally in both research and industry communities during the past decade, predominantly owing to the Big Data revolution. Artificial Intelligence (AI) and Machine Learning (ML) are bringing more complexities to data engineering applications, which are now integrated into data processing pipelines to process terabytes of data. Typically, a significant amount of time is spent on data preprocessing in these pipelines, and hence improving its efficiency directly impacts the overall pipeline performance. The community has recently embraced the concept of Dataframes as the de-facto data structure for data representation and manipulation. However, the most widely used serial Dataframes today (R, pandas) experience performance limitations while working on even moderately large data sets. We believe that there is plenty of room for improvement by taking a look at this problem from a high-performance computing point of view. In a prior publication, we presented a set of parallel processing patterns for distributed dataframe operators and the reference runtime implementation, Cylon. In this paper, we are expanding on the initial concept by introducing a cost model for evaluating the said patterns. Furthermore, we evaluate the performance of Cylon on the ORNL Summit supercomputer.

97 MATHEMATICS AND COMPUTING↗

Functional groups show distinct differences in nitrogen cycling during early stand development: implications for forest management.

Nutrient acquisition of forest stands is controlled by soil resource availability and belowground production, but tree species are rarely compared in this regard. Here, we examine ecological and management implications of nitrogen (N) dynamics during early forest stand development in productive commercial tree species with narrow (Populus deltoides Bartr. and Platanus occidentalis L.) and broad (Liquidambar styraciflua L. and Pinus taeda L.) site requirements while grown with a range of nutrient and water resources. We constructed N budgets by measuring N concentration ([N]) and N content (N{sub C}) of above- and belowground perennial and ephemeral tissues, determined N uptake (N{sub UP}), and calculated N use efficiency (NUE). Forest stands regulated [N] within species-specific operating ranges without clear temporal or treatment patterns, thus demonstrating equilibrium between tissue [N] and biomass accumulation. Forest stand N{sub C} and N{sub UP} increased with stand development and paralleled treatment patterns of biomass accumulation, suggesting productivity is tightly linked to N{sub UP}. Inclusion of above- and belowground ephemeral tissue turnover in N{sub UP} calculations demonstrated that maximum N demand for narrow-sites adapted species exceeded 200 kg N ha{sup -1} year{sup -1} while demand for broad-site adapted species was below this level. NUE was species dependent but not consistently influenced by N availability, suggesting relationships between NUE and resource availability were species dependent. Based on early stand development, species with broad site adaptability are favored for woody cropping systems because they maintain high above- and belowground productivity with minimal fertilization requirements due to higher NUE than narrow site adapted species.

Nitrogen use efficiency↗

Graphics processing unit accelerated three-dimensional model for the simulation of pulsed low-temperature plasmas

A 3-dimensional particle-in-cell/Monte Carlo collision simulation that is fully implemented on a graphics processing unit (GPU) is described and used to determine low-temperature plasma characteristics at high reduced electric field, E/n, in nitrogen gas. Details of implementation on the GPU using the NVIDIA Compute Unified Device Architecture framework are discussed with respect to efficient code execution. The software is capable of tracking around 10 × 10{sup 6} particles with dynamic weighting and a total mesh size larger than 10{sup 8} cells. Verification of the simulation is performed by comparing the electron energy distribution function and plasma transport parameters to known Boltzmann Equation (BE) solvers. Under the assumption of a uniform electric field and neglecting the build-up of positive ion space charge, the simulation agrees well with the BE solvers. The model is utilized to calculate plasma characteristics of a pulsed, parallel plate discharge. A photoionization model provides the simulation with additional electrons after the initial seeded electron density has drifted towards the anode. Comparison of the performance benefits between the GPU-implementation versus a CPU-implementation is considered, and a speed-up factor of 13 for a 3D relaxation Poisson solver is obtained. Furthermore, a factor 60 speed-up is realized for parallelization of the electron processes.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Efficient track-before detect algorithm with minimal prior knowledge

Apparatus, system and method for tracking an image target in a system, wherein a system receives an image comprising a plurality of pixels. The received image is processed via a plurality of different recursive motion model kernels in parallel to provide a plurality of kernel outputs, wherein each of the motion model kernels may include a respective pixel mask. Per-pixel energy is estimated of at least some of the plurality of kernel outputs. Velocity of at least one of the image pixels may also be estimated by generating a directional energy vector for each motion model kernel. The per-pixel energy and velocity estimates are fused to produce a fused estimate representing at least some of the motion model kernels for the image.

42 ENGINEERING↗

Detective quantum efficiency of photon-counting x-ray detectors

Purpose: Single-photon-counting (SPC) x-ray imaging has the potential to improve image quality and enable novel energy-dependent imaging methods. Similar to conventional detectors, optimizing image SPC quality will require systems that produce the highest possible detective quantum efficiency (DQE). This paper builds on the cascaded-systems analysis (CSA) framework to develop a comprehensive description of the DQE of SPC detectors that implement adaptive binning. Methods: The DQE of SPC systems can be described using the CSA approach by propagating the probability density function (PDF) of the number of image-forming quanta through simple quantum processes. New relationships are developed to describe PDF transfer through serial and parallel cascades to accommodate scatter reabsorption. Results are applied to hypothetical silicon and selenium-based flat-panel SPC detectors including the effects of reabsorption of characteristic/scatter photons from photoelectric and Compton interactions, stochastic conversion of x-ray energy to secondary quanta, depth-dependent charge collection, and electronic noise. Results are compared with a Monte Carlo study. Results: Depth-dependent collection efficiency can result in substantial broadening of photopeaks that in turn may result in reduced DQE at lower x-ray energies (20–45 keV). Double-counting interaction events caused by reabsorption of characteristic/scatter photons may result in falsely inflated image signal-to-noise ratio and potential overestimation of the DQE. Conclusions: The CSA approach is extended to describe signal and noise propagation through photoelectric and Compton interactions in SPC detectors, including the effects of escape and reabsorption of emission/scatter photons. High-performance SPC systems can be achieved but only for certain combinations of secondary conversion gain, depth-dependent collection efficiency, electronic noise, and reabsorption characteristics.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

RADAR: Runtime Asymmetric Data-access Driven Scientific Data Replication

Efficient I/O on large-scale spatiotemporal scientific data requires scrutiny of both the logical layout of the data (e.g., row-major vs. column-major) and the physical layout (e.g., distribution on parallel filesystems). For increasingly complex datasets, hand optimization is a difficult matter prone to error and not scalable to the increasing heterogeneity of analysis workloads. Given these factors, we present a partial data replication system called RADAR. We capture datatype- and collective-aware I/O access patterns (indicating logical access) via MPI-IO tracing and use a combination of coarse-grained and fine-grained performance modeling to evaluate and select optimized physical data distributions for the task at hand. Unlike conventional methods, we store all replica data and metadata, along with the original untouched data, under a single file container using the object abstraction in parallel filesystems. Our system results in manyfold improvements in some commonly used subvolume decomposition access patterns.Moreover, the modeling approach can determine whether such optimizations should be undertaken in the first place.

access pattern↗

Design Principles for Sparse Matrix Multiplication on the GPU

We implement two novel algorithms for sparse-matrix dense-matrix multiplication (SpMM) on the GPU. Our algorithms expect the sparse input in the popular compressed-sparse-row (CSR) format and thus do not require expensive format conversion. While previous SpMM work concentrates on thread-level parallelism, we additionally focus on latency hiding with instruction-level parallelism and load-balancing. We show, both theoretically and experimentally, that the proposed SpMM is a better fit for the GPU than previous approaches. We identify a key memory access pattern that allows efficient access into both input and output matrices that is crucial to getting excellent performance on SpMM. By combining these two ingredients---(i) merge-based load-balancing and (ii) row-major coalesced memory access---we demonstrate a 4.1x peak speedup and a 31.7% geomean speedup over state-of-the-art SpMM implementations on real-world datasets.

97 MATHEMATICS AND COMPUTING↗

Synchrotron X-ray topographic study on nature of threading mixed dislocations in 4H–SiC crystals grown by PVT method

Synchrotron X-ray Topography is a powerful technique to study defects structures particularly dislocation configurations in single crystals. Complementing this technique with geometrical and contrast analysis can enhance the efficiency of quantitatively characterizing defects. In this study, the use of Synchrotron White Beam X-ray Topography (SWBXT) to determine the line directions of threading dislocations in 4H–SiC axial slices (sample cut parallel to the growth axis from the boule) is demonstrated. This technique is based on the fact that the projected line directions of dislocations on different reflections are different. Another technique also discussed is the determination of the absolute Burgers vectors of threading mixed dislocations (TMDs) using Synchrotron Monochromatic Beam X-ray Topography (SMBXT). This technique utilizes the fact that the contrast from TMDs varies on SMBXT images as their Burgers vectors change. By comparing observed contrast with the contrast from threading dislocations provided by Ray Tracing Simulations, the Burgers vectors can be determined. Thereafter the distribution of TMDs with different Burgers vectors across the wafer is mapped and investigated.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Structure, growth and properties of a novel polar material, KSr{sub 4}B{sub 3}O{sub 9}

A novel polar material, KSr{sub 4}B{sub 3}O{sub 9}, has been obtained by the high temperature solution method for the first time. It crystallizes in the polar space group Ama2. The material exhibits a three-dimensional structure consisting of interconnecting KO{sub 10} groups, SrO{sub x} (x=8, 9) units and isolated BO{sub 3} triangles. DSC/TG curve shows that KSr{sub 4}B{sub 3}O{sub 9} melts incongruently. The calculated band structures and the density of states of KSr{sub 4}B{sub 3}O{sub 9} suggest that its direct gap is 3.897 eV. The absorption spectrum indicates the absorption edge of KSr{sub 4}B{sub 3}O{sub 9} is about 302 nm. It produces SHG intensity about as large as that of KH{sub 2}PO{sub 4} (KDP) and is phase matchable. - Graphical abstract: Viewed along the c-axis, B(2)O{sub 3} triangles and B(3)O{sub 3} triangles are parallel with the (0 1 0) face and opposite to each other. However, all the B(1)O{sub 3} triangles have the same direction along the c-axis. So the main SHG efficiency of the compound comes from B(1)O{sub 3} triangles according to the anion group theory. Highlights: Black-Right-Pointing-Pointer A polar material, KSr{sub 4}B{sub 3}O{sub 9}, has been obtained by high temperature solution method. Black-Right-Pointing-Pointer The material consists of KO{sub 10} units, SrO{sub x}(x=8, 9) units and isolated BO{sub 3} units. Black-Right-Pointing-Pointer KSr{sub 4}B{sub 3}O{sub 9} produces SHG intensity as large as that of KDP and is phase matchable. Black-Right-Pointing-Pointer The absorption edge of KSr{sub 4}B{sub 3}O{sub 9} is about 302 nm.

36 MATERIALS SCIENCE↗

A modified Bitter-type electromagnet and control system for cold atom experiments

We present a modified Bitter-type electromagnet which features high magnetic field, fine electronic properties and efficient heat removal. The electromagnet is constructed from a stack of copper layers separated by mica layers that have the same shape. A distinctive design of cooling channels on the insulating layers and the parallel ducts between the layers ensures low resistance for cooling water to flow. A continuous current control system is also made to regulate the current through the electromagnet. In our experiment, versatile electromagnets are applied to generate magnetic field and gradient field. From our measurements, a peak magnetic field of 1000 G and a peak gradient field of 80 G/cm are generated in the center of the apparatuses which are 7 cm and 5 cm away from the edge of each electromagnet with a current of 230 A and 120 A, respectively. With the effective feedback design in the current control system and cooling water flow of 3.8 l/min, the stability of the current through the electromagnets can reach 10{sup −5}.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Scaled experiments on cavity confined explosions in limestone and poly(methyl methacrylate)

A scaled experiment comprising a laser-driven explosion in a cavity is used to characterize the coupling of mechanical energy into the surrounding solid material. Experiments are performed using poly(methyl methacrylate) and dry Salem limestone as the explosion containment blocks materials, in which are milled scaled spherical cavities of various dimensions. Measurements of the coupled shock, taken with fiber optic probes at the cavity wall, show the critical radius where wall deformation transitions from plastic to elastic deformation. These measurements also provide a diagnostic of the air blast, which is validated against GEODYN simulation code. The measurement of the coupled shock amplitude taken farther from the wall in the linear region indicates increased coupling efficiency in small cavities over the range of scaled cavity radii from 6 to 20 m/kt 1/3 , a phenomenon not previously observed in experiments. A comparison of results taken in this experiment with a parallel experiment using high explosive (HE) as the source shows that coupled shocks generated with HE are characteristically different with much larger amplitude than those produced by a high energy density laser-driven source with the same yield. This experimental technique potentially provides a rapid and cost-effective method to analyze the consequences of a full-scale, low yield, buried explosion.

58 GEOSCIENCES↗

MAGNETOHYDRODYNAMIC SIMULATIONS OF RECONNECTION AND PARTICLE ACCELERATION: THREE-DIMENSIONAL EFFECTS

Magnetic fields can change their topology through a process known as magnetic reconnection. This process in not only important for understanding the origin and evolution of the large-scale magnetic field, but is seen as a possibly efficient particle accelerator producing cosmic rays mainly through the first-order Fermi process. In this work we study the properties of particle acceleration inserted in reconnection zones and show that the velocity component parallel to the magnetic field of test particles inserted in magnetohydrodynamic (MHD) domains of reconnection without including kinetic effects, such as pressure anisotropy, the Hall term, or anomalous effects, increases exponentially. Also, the acceleration of the perpendicular component is always possible in such models. We find that within contracting magnetic islands or current sheets the particles accelerate predominantly through the first-order Fermi process, as previously described, while outside the current sheets and islands the particles experience mostly drift acceleration due to magnetic field gradients. Considering two-dimensional MHD models without a guide field, we find that the parallel acceleration stops at some level. This saturation effect is, however, removed in the presence of an out-of-plane guide field or in three-dimensional models. Therefore, we stress the importance of the guide field and fully three-dimensional studies for a complete understanding of the process of particle acceleration in astrophysical reconnection environments.

79 ASTRONOMY AND ASTROPHYSICS↗

ON THE COOLING TAILS OF THERMONUCLEAR X-RAY BURSTS: THE IGR J17480-2446 LINK

The neutron star transient and 11 Hz X-ray pulsar IGR J17480-2446, recently discovered in the globular cluster Terzan 5, showed unprecedented bursting activity during its 2010 October-November outburst. We analyzed all X-ray bursts detected with the Rossi X-ray Timing Explorer and find strong evidence that they all have a thermonuclear origin, despite the fact that many do not show the canonical spectral softening along the decay imprinted on type I X-ray bursts by the cooling of the neutron star photosphere. We show that the persistent-to-burst power ratio is fully consistent with the accretion-to-thermonuclear efficiency ratio along the whole outburst, as is typical for type I X-ray bursts. The burst energy, peak luminosity, and daily-averaged spectral profiles all evolve smoothly throughout the outburst, in parallel with the persistent (non-burst) luminosity. We also find that the peak-burst to persistent luminosity ratio determines whether or not cooling is present in the bursts from IGR J17480-2446, and argue that the apparent lack of cooling is due to the 'non-cooling' bursts having both a lower peak temperature and a higher non-burst (persistent) emission. We conclude that the detection of cooling along the decay is a sufficient, but not a necessary condition to identify an X-ray burst as thermonuclear. Finally, we compare these findings with X-ray bursts from other rapidly accreting neutron stars.

79 ASTRONOMY AND ASTROPHYSICS↗

srMO-BO-3GP: A sequential regularized multi-objective Bayesian optimization for constrained design applications using an uncertain Pareto classifier

Bayesian optimization (BO) is an efficient and flexible global optimization framework that is applicable to a very wide range of engineering applications. To leverage the capability of the classical BO, many extensions, including multi-objective, multi-fidelity, parallelization, and latent-variable modeling, have been proposed to address the limitations of the classical BO framework. In this work, we propose a novel multi-objective BO formalism, called srMO-BO-3GP, to solve multi-objective optimization problems in a sequential setting. Three different Gaussian processes (GPs) are stacked together, where each of the GPs is assigned with a different task. The first GP is used to approximate a single-objective computed from the multi-objective definition, the second GP is used to learn the unknown constraints, and the third one is used to learn the uncertain Pareto frontier. At each iteration, a multi-objective augmented Tchebycheff function is adopted to convert multi-objective to single-objective, where the regularization with a regularized ridge term is also introduced to smooth the single-objective function. Finally, we couple the third GP along with the classical BO framework to explore the convergence and diversity of the Pareto frontier by the acquisition function for exploitation and exploration. The proposed framework is demonstrated using several numerical benchmark functions, as well as a thermomechanical finite element model for flip-chip package design optimization.

42 ENGINEERING↗

Cache Locality Optimization for Recursive Programs

We present an approach to optimize the cache locality for recursive programs by dynamically splicing--recursively interleaving--the execution of distinct function invocations. By utilizing data effect annotations, we identify concurrency and data reuse opportunities across function invocations and interleave them to reduce reuse distance. We present algorithms that efficiently track effects in recursive programs, detect interference and dependencies, and interleave execution of function invocations using user-level (non-kernel) lightweight threads. To enable multi-core execution, a program is parallelized using a nested fork/join programming model. Our cache optimization strategy is designed to work in the context of a random work stealing scheduler. We present an implementation using the MIT Cilk framework that demonstrates significant improvements in sequential and parallel performance, competitive with a state-of-the-art compile-time optimizer for loop programs and a domain- specific optimizer for stencil programs.

Lifflander, Jonathan↗

Continuous Succinic Acid Production by Actinobacillus succinogenes on Xylose-Enriched Hydrolysate

Bio-manufacturing of high-value chemicals in parallel to renewable biofuels has the potential to dramatically improve the overall economic landscape of integrated lignocellulosic biorefineries. However, this will require the generation of carbohydrate streams from lignocellulose in a form suitable for efficient microbial conversion and downstream processing appropriate to the desired end use, making overall process development, along with selection of appropriate target molecules, crucial to the integrated biorefinery. Succinic acid (SA), a high-value target molecule, can be biologically produced from sugars and has the potential to serve as a platform chemical for various chemical and polymer applications. However, the feasibility of microbial SA production at industrially relevant productivities and yields from lignocellulosic biorefinery streams has not yet been reported.

09 BIOMASS FUELS↗

Selection of Shear Horizontal Wave Transducers for Robotic Nondestructive Inspection in Harsh Environments

Harsh environments and confined spaces require that nondestructive inspections be conducted with robotic systems. Ultrasonic guided waves are well suited for robotic systems because they can provide efficient volumetric coverage when inspecting for various types of damage, including cracks and corrosion. Shear horizontal guided waves are especially well suited for robotic inspection because they are sensitive to cracks oriented perpendicular or parallel to the wave propagation direction and can be generated with electromagnetic acoustic transducers (EMATs) and magnetostrictive transducers (MSTs). Both types of transducers are investigated for crack detection in a stainless steel plate. The MSTs require the robot to apply a compressive normal force that creates frictional force coupling. However, the coupling is observed to be very dependent upon surface roughness and surface debris. The EMATs are coupled through the Lorentz force and are thus noncontact, although they depend on the lift off between transducer and substrate. After comparing advantages and disadvantages of each transducer for robotic inspection the EMATs are selected for application to canisters that store used nuclear fuel.

Choi, Sungho↗

Conversion of Resonance Parameters Between Wigner-Eisenbud R-matrix and Transition Matrix Pole Representation

The goal of this research is to develop a numerical code for translation between the current standard, Wigner-Eisenbud R-matrix parameters and a novel pole-representation of cross sections. The method follows Ducru et al. and constructs the conversion between these parametric representations of the scattering and reaction data. This pole representation, which expresses cross sections as a sum of simple poles (and their attendant residues), is examined in the context of the advantages provided by this compact, computationally efficient representation. These advantages include the ability to doppler broaden cross sections on-the-fly as well as a lower memory footprint when generating cross sections, an attractive feature to those interested in parallel GPU based Monte Carlo. Some conversion methods between R-matrix parameters and S-matrix poles have relied on curve fitting techniques. It is unknown weather these numerical methods necessarily preserve certain underlying physics properties of the standard R-matrix cross section model (such as unitarity). The direct conversion studied here aim to preserve such properties. This knowledge is of interest for future work into the generation of resonance parameter sensitivities.

CLUTCH↗