Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed parallelization”

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 739 records · Page 41

DEVELOPMENT OF A NUCLEAR FUEL SAFEGUARDS VERIFICATION TECHNOLOGY FOR NEW FACILITY TYPES: Use of fast neutron emission tomography for spent fuel verification

With the buildup of spent fuel inventories around the world from decommissioning activities, a tool for the efficient and effective verification of a variety of fuel types for long-term disposition is needed. Fast neutron emission tomography is a promising imaging technology that can address this problem and has been under development at Oak Ridge National Laboratory for the past two years. The goal of this capability is to detect the removal or substitution of individual fuel pins in spent nuclear fuel assemblies for international safeguards applications. The present imaging technique is based on neutron emissions originating primarily from curium-244, which is produced predominantly at the end of the exposure cycle. As a result, this technique may be sensitive not only to fuel pins that are removed or substituted subsequent to all irradiation, but also to fuel pins that are substituted and subsequently irradiated. At present, a laboratory prototype imager is under construction. The purpose of the prototype is to demonstrate an imaging capabilitysufficient to resolve individual fuel pins using commercially available boron straw detectors that can withstand the high gamma ray dose from a typical spent fuel assembly. To achieve both practical measurement durations and sufficient resolution to discern individual pins, the prototype imager employs a novel collimator design where the slits of a parallel-slit collimator are rearranged to distribute them around a ring surrounding the target fuel assembly. This paper will report on simulations and experimental results undertaken to converge on a final design, image reconstruction of sample fuel types of interest, and challenges associated.

Iyengar, Anagha↗

Modeling Passive Fast-Neutron Emission Tomography of Spent Nuclear Fuel

For the past year Oak Ridge National Laboratory has been developing a passive fast-neutron emission tomography capability. Like the passive gamma emission tomography capability presently being evaluated by the International Atomic Energy Agency, the goal of this capability is to detect the removal or substitution of individual fuel pins in spent nuclear fuel assemblies for international safeguards applications, such as verifying the integrity of an assembly before transfer to difficult-to-access storage. The present imaging technique is based on neutron emissions originating primarily from curium-244, which is produced predominantly at the end of the exposure cycle. As a result, this technique may be sensitive not only to fuel pins that are removed or substituted subsequent to all irradiation, but also to fuel pins that are substituted and subsequently irradiated. At present, a laboratory prototype imager is under construction. The purpose of the prototype is to demonstrate an imaging capability sufficient to resolve individual fuel pins using commercially available boron straw detectors that can withstand the high gamma ray dose from a typical spent fuel assembly. To achieve both practical measurement durations and sufficient resolution to discern individual pins, the prototype imager employs a novel collimator design where the slits of a parallel-slit collimator are rearranged to distribute them around a ring surrounding the target fuel assembly. This presentation will report on simulations to understand the neutron signal-to-noise ratio and gamma dose rate as a function of the collimator design, the neutron detector’s response to gamma rays as a function of dose rate, and image reconstruction challenges associated with this collimator design.

Hausladen, Paul↗

System-wide power management control via clock distribution network

An apparatus, method and computer program product for automatically controlling power dissipation of a parallel computing system that includes a plurality of processors. A computing device issues a command to the parallel computing system. A clock pulse-width modulator encodes the command in a system clock signal to be distributed to the plurality of processors. The plurality of processors in the parallel computing system receive the system clock signal including the encoded command, and adjusts power dissipation according to the encoded command.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Parallelization and checkpointing of GPU applications through program transformation

GPUs have emerged as a powerful tool for accelerating general-purpose applications. The availability of programming languages that makes writing general-purpose applications for running on GPUs tractable have consolidated GPUs as an alternative for accelerating general purpose applications. Among the areas that have benefited from GPU acceleration are: signal and image processing, computational fluid dynamics, quantum chemistry, and, in general, the High Performance Computing (HPC) Industry. In order to continue to exploit higher levels of parallelism with GPUs, multi-GPU systems are gaining popularity. In this context, single-GPU applications are parallelized for running in multi-GPU systems. Furthermore, multi-GPU systems help to solve the GPU memory limitation for applications with large application memory footprint. Parallelizing single-GPU applications has been approached by libraries that distribute the workload at runtime, however, they impose execution overhead and are not portable. On the other hand, on traditional CPU systems, parallelization has been approached through application transformation at pre-compile time, which enhances the application to distribute the workload at application level and does not have the issues of library-based approaches. Hence, a parallelization scheme for GPU systems based on application transformation is needed. Like any computing engine of today, reliability is also a concern in GPUs. GPUs are vulnerable to transient and permanent failures. Current checkpoint/restart techniques are not suitable for systems with GPUs. Checkpointing for GPU systems present new and interesting challenges, primarily due to the natural differences imposed by the hardware design, the memory subsystem architecture, the massive number of threads, and the limited amount of synchronization among threads. Therefore, a checkpoint/restart technique suitable for GPU systems is needed. The goal of this work is to exploit higher levels of parallelism and to develop support for application-level fault tolerance in applications using multiple GPUs. Our techniques reduce the burden of enhancing single-GPU applications to support these features. To achieve our goal, this work designs and implements a framework for enhancing a single-GPU OpenCL application through application transformation.

97 MATHEMATICS AND COMPUTING↗

Computational Challenges for Power System Operation

As the power grid technology evolution and information technology revolution converge, power grids are witnessing a revolutionary transition, represented by emerging grid technologies and large scale deployment of new sensors and meters in networks. This transition brings opportunities, as well as computational challenges in the field of power grid analysis and operation. This paper presents some research outcomes in the areas of parallel state estimation using the preconditioned conjugated gradient method, parallel contingency analysis with a dynamic load balancing scheme and distributed system architecture. Based on this research, three types of computational challenges are identified: highly coupled applications, loosely coupled applications, and centralized and distributed applications. Recommendations for future work for power grid applications are also presented.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Lighting system with heat distribution face plate

Lighting systems having a light source and a thermal management system are provided. The thermal management system includes synthetic jet devices, a heat sink and a heat distribution face plate. The synthetic jet devices are arranged in parallel to one and other and are configured to actively cool the lighting system. The heat distribution face plate is configured to radially transfer heat from the light source into the ambient air.

42 ENGINEERING↗

High Level Synthesis of RDF Queries for Graph Analytics

In this paper we present a set of techniques that enable the synthesis of efficient custom accelerators for memory intensive, irregular applications. To address irregular applications challenges (large memory footprints, unpredictable fine- grained data accesses, and high synchronization intensity), and exploit their opportunities (thread level parallelism, memory level parallelism), we propose a novel accelerator design which take advantage of an adaptive and Distributed Controller (DC) architecture, and a Memory Interface (MI) that supports parallel memory subsystems. Among the multitude of algorithms that may benefit from our solution, we focus on the acceleration of graph analytics applications, and in particular, on the syn- thesis of SPARQL queries on Resource Description Framework (RDF) databases. We achieve this objective by incorporating the synthesis techniques into Bambu, an Open Source high- level synthesis tools, and interfacing the system with GEMS, the Graph database Engine for Multithreaded Systems. The front- end of GEMS generates optimized C implementations of the input queries, modeled as pattern matching routines, which are then automatically synthesized by Bambu. We validate our approach synthesizing several SPARQL queries from the Lehigh University Benchmark (LUBM).

HLS, FPGA, Convey, RDF, GEMS, Memory-intensive,↗

Tandem Differential Mobility Analyzer/Aerodynamic Particle Sizer (APS) Handbook

The tandem differential mobility analyzer (TDMA) is a single instrument that cycles through a series of complementary measurements of the physical properties of size-resolved submicron particles. In 2008, the TDMA was augmented through the addition of an aerodynamic particle sizer (APS), which extends the upper limit of the measured size distribution into the supermicron range. These two instruments are operated in parallel, but because they are controlled by a common computer and because the size distributions measured by the two are integrated in the produced datastreams, they are described together here. Throughout the day, the TDMA sequentially measures submicron aerosol size distributions and size-resolved hygroscopic growth distributions. More specifically, the instrument is operated as a scanning DMA to measure size distributions and as a TDMA to measure size-resolved hygroscopicity. A typical measurement sequence requires roughly 45 minutes. Each morning additional measurements are made of the relative humidity (RH) dependent hygroscopicity and temperature-dependent volatility of size-resolved particles. When the outside temperature and RH are within acceptable ranges, the hydration state of size-resolved particles is also characterized. The measured aerosol distributions complement the array of aerosol instruments in the Aerosol Observing System (AOS) and provide additional details of the light-scattering and cloud-nucleating characteristics of the aerosol.

47 OTHER INSTRUMENTATION↗

Spectral Schur complement techniques for symmetric eigenvalue problems

This paper presents a domain decomposition-type method for solving real symmetric (Hermitian) eigenvalue problems in which we seek all eigenpairs in an interval [α; β] or a few eigenpairs next to a given real shift. Here, a Newton-based scheme is described whereby the problem is converted to one that deals with the interface nodes of the computational domain. This approach relies on the fact that the inner solves related to each local subdomain are relatively inexpensive. This Newton scheme exploits spectral Schur complements, and these lead to so-called eigenbranches, which are rational functions whose roots are eigenvalues of the original matrix. Theoretical and practical aspects of domain decomposition techniques for computing eigenvalues and eigenvectors are discussed. A parallel implementation is presented and its performance on distributed computing environments is illustrated by means of a few numerical examples.

parallel computing↗

Computing contingency statistics in parallel.

Statistical analysis is typically used to reduce the dimensionality of and infer meaning from data. A key challenge of any statistical analysis package aimed at large-scale, distributed data is to address the orthogonal issues of parallel scalability and numerical stability. Many statistical techniques, e.g., descriptive statistics or principal component analysis, are based on moments and co-moments and, using robust online update formulas, can be computed in an embarrassingly parallel manner, amenable to a map-reduce style implementation. In this paper we focus on contingency tables, through which numerous derived statistics such as joint and marginal probability, point-wise mutual information, information entropy, and {chi}{sup 2} independence statistics can be directly obtained. However, contingency tables can become large as data size increases, requiring a correspondingly large amount of communication between processors. This potential increase in communication prevents optimal parallel speedup and is the main difference with moment-based statistics where the amount of inter-processor communication is independent of data size. Here we present the design trade-offs which we made to implement the computation of contingency tables in parallel.We also study the parallel speedup and scalability properties of our open source implementation. In particular, we observe optimal speed-up and scalability when the contingency statistics are used in their appropriate context, namely, when the data input is not quasi-diffuse.

99 GENERAL AND MISCELLANEOUS↗

Summer Proceedings 2016: The Center for Computing Research at Sandia National Laboratories

Solving sparse linear systems from the discretization of elliptic partial differential equations (PDEs) is an important building block in many engineering applications. Sparse direct solvers can solve general linear systems, but are usually slower and use much more memory than effective iterative solvers. To overcome these two disadvantages, a hierarchical solver (LoRaSp) based on H2-matrices was introduced in [22]. Here, we have developed a parallel version of the algorithm in LoRaSp to solve large sparse matrices on distributed memory machines. On a single processor, the factorization time of our parallel solver scales almost linearly with the problem size for three-dimensional problems, as opposed to the quadratic scalability of many existing sparse direct solvers. Moreover, our solver leads to almost constant numbers of iterations, when used as a preconditioner for Poisson problems. On more than one processor, our algorithm has significant speedups compared to sequential runs. With this parallel algorithm, we are able to solve large problems much faster than many existing packages as demonstrated by the numerical experiments.

97 MATHEMATICS AND COMPUTING↗

Dosimetric properties of a proton beamline dedicated to the treatment of ocular disease

Purpose: A commercial proton eyeline has been developed to treat ocular disease. Radiotherapy of intraocular lesions (e.g., uveal melanoma, age-related macular degeneration) requires sharp dose gradients to avoid critical structures like the macula and optic disc. A high dose rate is needed to limit patient gazing times during delivery of large fractional dose. Dose delivery needs to be accurate and predictable, not in the least because current treatment planning algorithms have limited dose modeling capabilities. The purpose of this paper is to determine the dosimetric properties of a new proton eyeline. These properties are compared to those of existing systems and evaluated in the context of the specific clinical requirements of ocular treatments. Methods: The eyeline is part of a high-energy, cyclotron-based proton therapy system. The energy at the entrance of the eyeline is 105 MeV. A range modulator (RM) wheel generates the spread-out Bragg peak, while a variable range shifter system adjusts the range and spreads the beam laterally. The range can be adjusted from 0.5 up to 3.4 g/cm{sup 2}; the modulation width can be varied in steps of 0.3 g/cm{sup 2} or less. Maximum field diameter is 2.5 cm. All fields can be delivered with a dose rate of 30 Gy/min or more. The eyeline is calibrated according to the IAEA TRS-398 protocol using a cylindrical ionization chamber. Depth dose distributions and dose/MU are measured with a parallel-plate ionization chamber; lateral profiles with radiochromic film. The dose/MU is modeled as a function of range, modulation width, and instantaneous MU rate with fit parameters determined per option (RM wheel). Results: The distal fall-off of the spread-out Bragg peak is 0.3 g/cm{sup 2}, larger than for most existing systems. The lateral penumbra varies between 0.9 and 1.4 mm, except for fully modulated fields that have a larger penumbra at skin. The source-to-axis distance is found to be 169 cm. The dose/MU shows a strong dependence on range (up to 4%/mm). A linear increase in dose/MU as a function of instantaneous MU rate is observed. The dose/MU model describes the measurements with an accuracy of ±2%. Neutron dose is found to be 146 ± 102 μSv/Gy at the contralateral eye and 19 ± 13 μSv/Gy at the chest. Conclusions: Measurements show the proton eyeline meets the requirements to effectively treat ocular disease.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

SU-E-T-257: Development of a New Endorectal Balloon with An Unfoldable Radiochromic Film for In-Vivo Rectal Dosimetry During Prostate Cancer Radiotherapy

Purpose: We developed an endorectal balloon for in-vivo rectal dosimetry in two-dimensions, and evaluated its dosimetric properties for the radiation treatment of prostate cancer. Methods: The endorectal balloon for in-vivo rectal dosimetry is equipped with a radiochromic film so that two-dimensional dose distribution can be measured in the rectal wall. The film is unrolled as the balloon is inflated, and it is rolled as the balloon is deflated. The outer diameter of the balloon is about 14 mm before inflating it, but its outer diameter can be increased up to about 50 mm after inflating it with 80 ml distilled water. The size of the film is 80(L) x 64(W) mm2, so large as to measure a dose distribution of an anterior half of the rectal wall. After it was inserted into a fabricated rectal phantom, the phantom was scanned by a CT scanner and 5 Gy was delivered to a target inside the phantom with a 15 MV photon beam in AP direction. Finally, the dose distribution measured in the endorectal balloon was compared with that of the treatment plan. Results: The two dose distributions were compared each other in the parallel and the perpendicular directions along an axis of the balloon. The two dose profiles analyzed from the radiochromic film agreed well with the plan within 3% for 15 MV photon beam. Conclusion: An endorectal balloon for two-dimensional in-vivo rectal dosimetry was developed and its dosimetric effectiveness was evaluated for the radiation treatment of prostate cancer. The measured dose distributions showed good agreement with the plans.

60 APPLIED LIFE SCIENCES↗

SALSA3D : a global 3D p-velocity model of the Earth's crust and mantle for improved event location.

To test the hypothesis that high quality 3D Earth models will produce seismic event locations which are more accurate and more precise, we are developing a global 3D P wave velocity model of the Earth's crust and mantle using seismic tomography. In this paper, we present the most recent version of our model, SALSA3D version 1.5, and demonstrate its ability to reduce mislocations for a large set of realizations derived from a carefully chosen set of globally-distributed ground truth events. Our model is derived from the latest version of the Ground Truth (GT) catalog of P and Pn travel time picks assembled by Los Alamos National Laboratory. To prevent over-weighting due to ray path redundancy and to reduce the computational burden, we cluster rays to produce representative rays. Reduction in the total number of ray paths is {approx}50%. The model is represented using the triangular tessellation system described by Ballard et al. (2009), which incorporates variable resolution in both the geographic and radial dimensions. For our starting model, we use a simplified two layer crustal model derived from the Crust 2.0 model over a uniform AK135 mantle. Sufficient damping is used to reduce velocity adjustments so that ray path changes between iterations are small. We obtain proper model smoothness by using progressive grid refinement, refining the grid only around areas with significant velocity changes from the starting model. At each grid refinement level except the last one we limit the number of iterations to prevent convergence thereby preserving aspects of broad features resolved at coarser resolutions. Our approach produces a smooth, multi-resolution model with node density appropriate to both ray coverage and the velocity gradients required by the data. This scheme is computationally expensive, so we use a distributed computing framework based on the Java Parallel Processing Framework, providing us with {approx}400 processors. Resolution of our model is assessed using a variation of the standard checkerboard method. We compare the travel-time prediction and location capabilities of SALSA3D to standard 1D models via location tests on a global event set with GT of 5 km or better. These events generally possess hundreds of Pn and P picks from which we generate different realizations of station distributions, yielding a range of azimuthal coverage and ratios of teleseismic to regional arrivals, with which we test the robustness and quality of relocation. The SALSA3D model reduces mislocation over standard 1D ak135 regardless of Pn to P ratio, with the improvement being most pronounced at higher azimuthal gaps.

99 GENERAL AND MISCELLANEOUS↗

A global 3D P-velocity model of the Earth's crust and mantle for improved event location : SALSA3D.

To test the hypothesis that high quality 3D Earth models will produce seismic event locations which are more accurate and more precise, we are developing a global 3D P wave velocity model of the Earth's crust and mantle using seismic tomography. In this paper, we present the most recent version of our model, SALSA3D version 1.5, and demonstrate its ability to reduce mislocations for a large set of realizations derived from a carefully chosen set of globally-distributed ground truth events. Our model is derived from the latest version of the Ground Truth (GT) catalog of P and Pn travel time picks assembled by Los Alamos National Laboratory. To prevent over-weighting due to ray path redundancy and to reduce the computational burden, we cluster rays to produce representative rays. Reduction in the total number of ray paths is {approx}50%. The model is represented using the triangular tessellation system described by Ballard et al. (2009), which incorporates variable resolution in both the geographic and radial dimensions. For our starting model, we use a simplified two layer crustal model derived from the Crust 2.0 model over a uniform AK135 mantle. Sufficient damping is used to reduce velocity adjustments so that ray path changes between iterations are small. We obtain proper model smoothness by using progressive grid refinement, refining the grid only around areas with significant velocity changes from the starting model. At each grid refinement level except the last one we limit the number of iterations to prevent convergence thereby preserving aspects of broad features resolved at coarser resolutions. Our approach produces a smooth, multi-resolution model with node density appropriate to both ray coverage and the velocity gradients required by the data. This scheme is computationally expensive, so we use a distributed computing framework based on the Java Parallel Processing Framework, providing us with {approx}400 processors. Resolution of our model is assessed using a variation of the standard checkerboard method. We compare the travel-time prediction and location capabilities of SALSA3D to standard 1D models via location tests on a global event set with GT of 5 km or better. These events generally possess hundreds of Pn and P picks from which we generate different realizations of station distributions, yielding a range of azimuthal coverage and ratios of teleseismic to regional arrivals, with which we test the robustness and quality of relocation. The SALSA3D model reduces mislocation over standard 1D ak135 regardless of Pn to P ratio, with the improvement being most pronounced at higher azimuthal gaps.

99 GENERAL AND MISCELLANEOUS↗

A global 3D P-velocity model of the Earth's crust and mantle for improved event location.

To test the hypothesis that high quality 3D Earth models will produce seismic event locations which are more accurate and more precise, we are developing a global 3D P wave velocity model of the Earth's crust and mantle using seismic tomography. In this paper, we present the most recent version of our model, SALSA3D (SAndia LoS Alamos) version 1.4, and demonstrate its ability to reduce mislocations for a large set of realizations derived from a carefully chosen set of globally-distributed ground truth events. Our model is derived from the latest version of the Ground Truth (GT) catalog of P and Pn travel time picks assembled by Los Alamos National Laboratory. To prevent over-weighting due to ray path redundancy and to reduce the computational burden, we cluster rays to produce representative rays. Reduction in the total number of ray paths is > 55%. The model is represented using the triangular tessellation system described by Ballard et al. (2009), which incorporates variable resolution in both the geographic and radial dimensions. For our starting model, we use a simplified two layer crustal model derived from the Crust 2.0 model over a uniform AK135 mantle. Sufficient damping is used to reduce velocity adjustments so that ray path changes between iterations are small. We obtain proper model smoothness by using progressive grid refinement, refining the grid only around areas with significant velocity changes from the starting model. At each grid refinement level except the last one we limit the number of iterations to prevent convergence thereby preserving aspects of broad features resolved at coarser resolutions. Our approach produces a smooth, multi-resolution model with node density appropriate to both ray coverage and the velocity gradients required by the data. This scheme is computationally expensive, so we use a distributed computing framework based on the Java Parallel Processing Framework, providing us with {approx}400 processors. Resolution of our model is assessed using a variation of the standard checkerboard method, as well as by directly estimating the diagonal of the model resolution matrix based on the technique developed by Bekas, et al. We compare the travel-time prediction and location capabilities of this model over standard 1D models. We perform location tests on a global, geographically-distributed event set with ground truth levels of 5 km or better. These events generally possess hundreds of Pn and P phases from which we can generate different realizations of station distributions, yielding a range of azimuthal coverage and proportions of teleseismic to regional arrivals, with which we test the robustness and quality of relocation. The SALSA3D model reduces mislocation over standard 1D ak135, especially with increasing azimuthal gap. The 3D model appears to perform better for locations based solely or dominantly on regional arrivals, which is not unexpected given that ak135 represents a global average and cannot therefore capture local and regional variations.

99 GENERAL AND MISCELLANEOUS↗

A global 3D P-Velocity model of the Earth%3CU%2B2019%3Es crust and mantle for improved event location.

To test the hypothesis that high quality 3D Earth models will produce seismic event locations which are more accurate and more precise, we are developing a global 3D P wave velocity model of the Earth's crust and mantle using seismic tomography. In this paper, we present the most recent version of our model, SALSA3D (SAndia LoS Alamos) version 1.4, and demonstrate its ability to reduce mislocations for a large set of realizations derived from a carefully chosen set of globally-distributed ground truth events. Our model is derived from the latest version of the Ground Truth (GT) catalog of P and Pn travel time picks assembled by Los Alamos National Laboratory. To prevent over-weighting due to ray path redundancy and to reduce the computational burden, we cluster rays to produce representative rays. Reduction in the total number of ray paths is > 55%. The model is represented using the triangular tessellation system described by Ballard et al. (2009), which incorporates variable resolution in both the geographic and radial dimensions. For our starting model, we use a simplified two layer crustal model derived from the Crust 2.0 model over a uniform AK135 mantle. Sufficient damping is used to reduce velocity adjustments so that ray path changes between iterations are small. We obtain proper model smoothness by using progressive grid refinement, refining the grid only around areas with significant velocity changes from the starting model. At each grid refinement level except the last one we limit the number of iterations to prevent convergence thereby preserving aspects of broad features resolved at coarser resolutions. Our approach produces a smooth, multi-resolution model with node density appropriate to both ray coverage and the velocity gradients required by the data. This scheme is computationally expensive, so we use a distributed computing framework based on the Java Parallel Processing Framework, providing us with {approx}400 processors. Resolution of our model is assessed using a variation of the standard checkerboard method, as well as by directly estimating the diagonal of the model resolution matrix based on the technique developed by Bekas, et al. We compare the travel-time prediction and location capabilities of this model over standard 1D models. We perform location tests on a global, geographically-distributed event set with ground truth levels of 5 km or better. These events generally possess hundreds of Pn and P phases from which we can generate different realizations of station distributions, yielding a range of azimuthal coverage and proportions of teleseismic to regional arrivals, with which we test the robustness and quality of relocation. The SALSA3D model reduces mislocation over standard 1D ak135, especially with increasing azimuthal gap. The 3D model appears to perform better for locations based solely or dominantly on regional arrivals, which is not unexpected given that ak135 represents a global average and cannot therefore capture local and regional variations.

99 GENERAL AND MISCELLANEOUS↗