Search NASA⌕ Search

SEARCH · Search NASA

Results for “graphical”

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 703 records · Page 39

Parallel Latent Semantic Analysis using a Graphics Processing Unit

Latent Semantic Analysis (LSA) can be used to reduce the dimensions of large Term-Document datasets using Singular Value Decomposition. However, with the ever expanding size of data sets, current implementations are not fast enough to quickly and easily compute the results on a standard PC. The Graphics Processing Unit (GPU) can solve some highly parallel problems much faster than the traditional sequential processor (CPU). Thus, a deployable system using a GPU to speedup large-scale LSA processes would be a much more effective choice (in terms of cost/performance ratio) than using a computer cluster. In this paper, we presented a parallel LSA implementation on the GPU, using NVIDIA Compute Unified Device Architecture (CUDA) and Compute Unified Basic Linear Algebra Subprograms (CUBLAS). The performance of this implementation is compared to traditional LSA implementation on CPU using an optimized Basic Linear Algebra Subprograms library. For large matrices that have dimensions divisible by 16, the GPU algorithm ran five to six times faster than the CPU version.

59 BASIC BIOLOGICAL SCIENCES↗

Beowulf - Beta-Gamma Detector Calibration Graphical User Interface

Pacific Northwest National Laboratory (PNNL) has demonstrated significant advancement in using beta-gamma coincidence detectors to detect a wide range of radioxenon isotopes. To obtain accurate activities with the detector it must be properly calibrated by measuring a series of calibration gas samples. The data is analyzed to create the calibration block used in the International Monitoring System file format. Doing the calibration manually has proven to be tedious and prone to errors, requiring a high degree of expertise. The Beowulf graphical user interface (GUI) is a software application that encompasses several components of the calibration task and generates a calibration block, as well as, a detailed report describing the specific calibration process used. This additional document can be used as a Quality assurance certificate to assist in auditing the calibration. This paper consists of two sections. Section 1 will describe the capabilities of Beowulf and section 2 will be a representative report generated or the 137Cs calibration and quality assurance source.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Massively Parallel Latent Semantic Analyzes using a Graphics Processing Unit

Latent Semantic Indexing (LSA) aims to reduce the dimensions of large Term-Document datasets using Singular Value Decomposition. However, with the ever expanding size of data sets, current implementations are not fast enough to quickly and easily compute the results on a standard PC. The Graphics Processing Unit (GPU) can solve some highly parallel problems much faster than the traditional sequential processor (CPU). Thus, a deployable system using a GPU to speedup large-scale LSA processes would be a much more effective choice (in terms of cost/performance ratio) than using a computer cluster. Due to the GPU s application-specific architecture, harnessing the GPU s computational prowess for LSA is a great challenge. We present a parallel LSA implementation on the GPU, using NVIDIA Compute Unified Device Architecture and Compute Unified Basic Linear Algebra Subprograms. The performance of this implementation is compared to traditional LSA implementation on CPU using an optimized Basic Linear Algebra Subprograms library. After implementation, we discovered that the GPU version of the algorithm was twice as fast for large matrices (1000x1000 and above) that had dimensions not divisible by 16. For large matrices that did have dimensions divisible by 16, the GPU algorithm ran five to six times faster than the CPU version. The large variation is due to architectural benefits the GPU has for matrices divisible by 16. It should be noted that the overall speeds for the CPU version did not vary from relative normal when the matrix dimensions were divisible by 16. Further research is needed in order to produce a fully implementable version of LSA. With that in mind, the research we presented shows that the GPU is a viable option for increasing the speed of LSA, in terms of cost/performance ratio.

99 GENERAL AND MISCELLANEOUS↗

Graphics Processing Unit Enhanced Parallel Document Flocking Clustering

Analyzing and clustering documents is a complex problem. One explored method of solving this problem borrows from nature, imitating the flocking behavior of birds. One limitation of this method of document clustering is its complexity O(n2). As the number of documents grows, it becomes increasingly difficult to generate results in a reasonable amount of time. In the last few years, the graphics processing unit (GPU) has received attention for its ability to solve highly-parallel and semi-parallel problems much faster than the traditional sequential processor. In this paper, we have conducted research to exploit this archi- tecture and apply its strengths to the flocking based document clustering problem. Using the CUDA platform from NVIDIA, we developed a doc- ument flocking implementation to be run on the NVIDIA GEFORCE GPU. Performance gains ranged from thirty-six to nearly sixty times improvement of the GPU over the CPU implementation.

Cui, Xiaohui↗

Digital Data Acquisition Graphical User Interface

Traditional radioxenon measurements have been done by ground based fixed systems, however in recent years there has been an increased need for systems capable of quick deployment or even complete mobility. Using the Pixie-4 data acquisition (DAQ) system can help reduce the electronics footprint of both current systems, like the radioxenon Radionuclide Laboratory 16 (RL-16) and the Swedish Automatic Unit for Noble Gas Acquisition (SAUNA), as well as future systems. Pacific Northwest National Laboratory (PNNL) has developed a Linux based Nyx graphical user interface (GUI) for Pixie-4 cards. The Nyx software can be installed on various Linux platforms and is written in C++. This software offers a rich user interface for configuring and operating the Pixie4 card and PNNL designed high voltage (HV) cards. Nyx allows one to quickly get a nuclear detector operational by maintaining the core diagnostic features built into the Pixie-4 cards. First, Nyx maintains the multitude of adjustable parameters accessible in the Pixie-4 cards, which allows one to customize settings to take full advantage of a particular detector. Nyx also maintains an oscilloscope feature which is extremely useful to optimize settings and to verify proper detector behavior and is often the first feature used in Nyx during detector setup. Finally, Nyx allows the user to collect data in several formats including full pulse shapes to basic histograms. Overall, it is the corner stone for the transition of beta-gamma systems to a state-of-the-art digitizing DAQ.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Graphically Contracted Function Construction with the Recursive Pairwise Merge Algorithm

A new implementation of a recursive pairwise merge algorithm to construct a GCF from a list of CSF expansion coefficients is presented. The essential new feature is the preallocation of some work arrays used within the intermediate steps of the merge procedure. This results in roughly an order of magnitude improvement in overall efficiency and also approximately eliminates a factor of n, the molecular orbital dimension, from the original implementation. Initial application of this merge procedure to a series of H m molecules shows that the GCF wave functions can be represented well both with delocalized canonical Hartree-Fock orbitals and with localized molecular orbitals. Finally, for a given wave function complexity, as measured by the average facet count, $\bar{\text {f}}$, the delocalized Hartree-Fock orbitals show smaller errors for small $\bar{\text {f}}$ values, while the localized orbitals show smaller errors for larger $\bar{\text {f}}$ values.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Topology Estimation using Graphical Models in Multi-Phase Power Distribution Grids

Power distribution grids are structurally operated radially, such that energized lines form a collection of trees with a substation at the root of each tree. The operational topology may change from time to time, however tracking these changes, even though important for the distribution grid operation and control, is hindered by limited real-time monitoring. This work develops a learning framework to reconstruct the radial operational structure of the distribution grid from synchronized voltage measurements. To detect operational lines our learning algorithm uses conditional independence tests for continuous random variables that is applicable to a wide class of probability distributions and Gaussian injections in particular. Furthermore, our algorithm applies to the practical case of unbalanced three-phase power flow. The performance is validated on AC power flow simulations over three phase IEEE distribution grid test cases.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Large-scale Nanostructure Simulations from X-ray Scattering Data On Graphics Processor Clusters

X-ray scattering is a valuable tool for measuring the structural properties of materialsused in the design and fabrication of energy-relevant nanodevices (e.g., photovoltaic, energy storage, battery, fuel, and carbon capture andsequestration devices) that are key to the reduction of carbon emissions. Although today's ultra-fast X-ray scattering detectors can provide tremendousinformation on the structural properties of materials, a primary challenge remains in the analyses of the resulting data. We are developing novelhigh-performance computing algorithms, codes, and software tools for the analyses of X-ray scattering data. In this paper we describe two such HPCalgorithm advances. Firstly, we have implemented a flexible and highly efficient Grazing Incidence Small Angle Scattering (GISAXS) simulation code based on theDistorted Wave Born Approximation (DWBA) theory with C++/CUDA/MPI on a cluster of GPUs. Our code can compute the scattered light intensity from any givensample in all directions of space; thus allowing full construction of the GISAXS pattern. Preliminary tests on a single GPU show speedups over 125x compared tothe sequential code, and almost linear speedup when executing across a GPU cluster with 42 nodes, resulting in an additional 40x speedup compared to usingone GPU node. Secondly, for the structural fitting problems in inverse modeling, we have implemented a Reverse Monte Carlo simulation algorithm with C++/CUDAusing one GPU. Since there are large numbers of parameters for fitting in the in X-ray scattering simulation model, the earlier single CPU code required weeks ofruntime. Deploying the AccelerEyes Jacket/Matlab wrapper to use GPU gave around 100x speedup over the pure CPU code. Our further C++/CUDA optimization deliveredan additional 9x speedup.

36 MATERIALS SCIENCE↗

Development of Improved Graphical Displays for an Advanced Outage Control Center, Employing Human Factors Principles for Outage Schedule Management

The long-term viability of existing nuclear power plants in the United States (U.S.) is dependent upon a number of factors, including maintaining high capacity factors, maintaining nuclear safety, and reducing operating costs, particularly those associated with refueling outages. Refueling outages typically take 20-30 days, and for existing light water NPPs in the U.S., the reactor cannot be in operation during the outage. Furthermore, given that many NPPs generate between $1-1.5 million/day in revenue when in operation, there is considerable interest in shortening the length of refueling outages. Yet refueling outages are highly complex operations, involving multiple concurrent and dependent activities that are somewhat challenging to coordinate; therefore, finding ways to improve refueling outage performance, while maintaining nuclear safety has proven to be difficult. The Advanced Outage Control Center (AOCC) project is a research and development (R&D) demonstration activity under the LWRS Program. LWRS is an R&D program that works closely with industry R&D programs to establish technical foundations for the licensing and managing of long-term, safe, and economical operation of current fleet of NPPs. As such, the LWRS Advanced Outage Control Center project has the goal of improving the management of commercial NPP refueling outages. To accomplish this goal, INL is developing an advanced outage control center (OCC) that is specifically designed to maximize the usefulness of communication and collaboration technologies for outage coordination and problem resolution activities. The overall focus is on developing an AOCC with the following capabilities that enables plant and OCC staff to; Collaborate in real-time to address emergent issues; Effectively communicate outage status to all workers involved in the outage; Effectively communicate discovered conditions in the field to the OCC; Provide real-time work status; Provide automatic pending support notifications; Provide real-time requirements monitoring; Maximize their collective situational awareness to improve decision-making; and Leverage macro data to better support resource allocation. INL has partnered with several commercial NPP utilities to develop a number of advanced outage management technologies. These outage management technologies have focused on both collaborative technologies for control centers and developing mobile technologies for NPP field workers. This report describes recent efforts made in developing a suite of outage technologies to support more effective schedule management. Currently, a master outage schedule is created months in advance using the plant’s existing scheduling software (e.g., Primavera P6). Typically, during the outage, the latest version of the schedule is printed at the beginning of each shift. INL and its partners are developing technologies that will have capabilities such as Automatic Schedule Updating, Automatic Pending Support Notifications, and the ability to allocate and schedule outage support task resources on a sub-hour basis (e.g., outage Micro-Scheduling). The remaining sections of this report describe in more detail the scheduling challenges that occur during outages, how the outage scheduling technologies INL is developing helps address those challenges, and the latest developments on this task (e.g., work accomplished to date and the path forward)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Low Lunar Orbit Design via Graphical Manipulation of Eccentricity Vector Evolution

Low lunar orbits, such as those used by GRAIL and LRO, experience predictable variations in the evolution of their eccentricity vectors. These variations are nearly invariant with respect to the initial eccentricity and argument of periapse and change only in the details with respect to the initial semi-major axis. These properties suggest that manipulating the eccentricity vector evolution directly can give insight into orbit maintenance designs and can reduce the number of propagations required. A trio of techniques for determining the desired maneuvers is presented in the context of the GRAIL extended mission.

techniques↗

Automatic Construction of Anomaly Detectors from Graphical Models

Detection of rare or previously unseen attacks in cyber security presents a central challenge: how does one search for a sufficiently wide variety of types of anomalies and yet allow the process to scale to increasingly complex data? In particular, creating each anomaly detector manually and training each one separately presents untenable strains on both human and computer resources. In this paper we propose a systematic method for constructing a potentially very large number of complementary anomaly detectors from a single probabilistic model of the data. Only one model needs to be trained, but numerous detectors can then be implemented. This approach promises to scale better than manual methods to the complex heterogeneity of real-life data. As an example, we develop a Latent Dirichlet Allocation probability model of TCP connections entering Oak Ridge National Laboratory. We show that several detectors can be automatically constructed from the model and will provide anomaly detection at flow, sub-flow, and host (both server and client) levels. This demonstrates how the fundamental connection between anomaly detection and probabilistic modeling can be exploited to develop more robust operational solutions.

anomaly detection↗