Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed computation”

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 109 records · Page 6

Distributed computer system enhances productivity for SRB joint optimization

The Programming System for Structural Synthesis software package, which couples structural analysis and optimization, has been distributed over a network of work stations for use in Space Shuttle Solid Rocket Booster joint redesign optimization. Finite difference computing techniques were applied to the optimization gradients in parallel execution, allowing several work stations to simultaneously contribute to the problem's solution. An optimal joint shape was obtained which achieves minimum weight while keeping the gap between joints well closed and limiting structural stresses. The optimization cycle was reduced from two hours to one-half hour.

Rogers, James L., Jr.↗

Redundancy management for efficient fault recovery in NASA's distributed computing system

The management of redundancy in computer systems was studied and guidelines were provided for the development of NASA's fault-tolerant distributed systems. Fault recovery and reconfiguration mechanisms were examined. A theoretical foundation was laid for redundancy management by efficient reconfiguration methods and algorithmic diversity. Algorithms were developed to optimize the resources for embedding of computational graphs of tasks in the system architecture and reconfiguration of these tasks after a failure has occurred. The computational structure represented by a path and the complete binary tree was considered and the mesh and hypercube architectures were targeted for their embeddings. The innovative concept of Hybrid Algorithm Technique was introduced. This new technique provides a mechanism for obtaining fault tolerance while exhibiting improved performance.

Malek, Miroslaw↗

Using the Global Arrays Toolkit to Reimplement NumPy for Distributed Computation

Global Arrays (GA) is a software system from Pacific Northwest National Laboratory that enables an efficient, portable, and parallel shared memory programming interface to manipulate distributed defense arrays. Using a combination of GA and NumPy, we have reimplemented NumPy as a distributed drop-in replacement called Global Arrays in NumPy (GAiN). Scalability studies will be presented showing the utility of developing serial NumPy codes which can later run on more capable clusters or supercomputers.

Global Arrays↗

Estimation of Errors in Stress Distributions Computed in Finite Element Simulations of Polycrystals

The accuracy of the stresses predicted from crystal plasticity-based finite element formulation depends on estimation and control of the errors associated with the discretization. In the current work, the errors in the stress distribution are estimated in virtual polycrystalline samples of α-phase titanium (hexagonal close-packed phase of Ti–6Al– 4 V). To estimate the error, the stress field, which does not possess inter-element continuity, is smoothed over a grain using an L 2 projection, thereby providing continuous stress distributions with inter-element continuity. The differences between the continuous (smooth) and discontinuous (raw) stress fields are calculated at individual Gauss quadrature points and used to estimate errors for corresponding elements and grains. Error estimations are performed for a Voronoi-tessellated microstructure, an equiaxed microstructure, and two microstructures with varying grain sizes for tensile loading extending into the fully plastic regime (≈ 5% extension). Magnitudes of the errors are found to depend on microstructural characteristics, particularly the shape and size of grains. Samples having variations in grain size or having less spherical grains exhibited higher errors than samples with uniformly sized, equiaxed grains, with the size variations having a more pronounced effect. In conclusion, errors correlate with proximity to grain boundaries at small (elastic) strains and with deformation-induced features (deformation bands) at large strains.

42 ENGINEERING↗

Intercommunications in Real Time, Redundant, Distributed Computer System

An investigation into the applicability of fiber optic communication techniques to real time avionic control systems, in particular the total automatic flight control system used for the VSTOL aircraft is presented. The system consists of spatially distributed microprocessors. The overall control function is partitioned to yield a unidirectional data flow between the processing elements (PE). System reliability is enhanced by the use of triple redundancy. Some general overall system specifications are listed here to provide the necessary background for the requirements of the communications system.

Zanger, H.↗

Seamless integration of commercial Clouds with ATLAS Distributed Computing

The CERN ATLAS Experiment successfully uses a worldwide dis-tributed computing Grid infrastructure to support its physics programme at the Large Hadron Collider (LHC). The Grid workflow system PanDA routinely manages up to 700,000 concurrently running production and analysis jobs to process simulation and detector data. In total more than 500 PB of data are distributed over more than 150 sites in the WLCG and handled by the ATLAS data management system Rucio. To prepare for the ever growing data rate in future LHC runs new developments are underway to embrace industry accepted protocols and technologies, and utilize opportunistic resources in a standard way. This paper reviews how the Google and Amazon Cloud computing ser-vices have been seamlessly integrated as a Grid site within PanDA and Rucio. Performance and brief cost evaluations will be discussed. Such setups could offer advanced Cloud tool-sets and provide added value for analysis facilities that are under discussions for LHC Run-4.

97 MATHEMATICS AND COMPUTING↗

Framework and Method for Controlling a Robotic System Using a Distributed Computer Network

A robotic system for performing an autonomous task includes a humanoid robot having a plurality of compliant robotic joints, actuators, and other integrated system devices that are controllable in response to control data from various control points, and having sensors for measuring feedback data at the control points. The system includes a multi-level distributed control framework (DCF) for controlling the integrated system components over multiple high-speed communication networks. The DCF has a plurality of first controllers each embedded in a respective one of the integrated system components, e.g., the robotic joints, a second controller coordinating the components via the first controllers, and a third controller for transmitting a signal commanding performance of the autonomous task to the second controller. The DCF virtually centralizes all of the control data and the feedback data in a single location to facilitate control of the robot across the multiple communication networks.

Sanders, Adam M.↗

System analysis for the Huntsville Operation Support Center distributed computer system

A simulation model of the NASA Huntsville Operational Support Center (HOSC) was developed. This simulation model emulates the HYPERchannel Local Area Network (LAN) that ties together the various computers of HOSC. The HOSC system is a large installation of mainframe computers such as the Perkin Elmer 3200 series and the Dec VAX series. A series of six simulation exercises of the HOSC model is described using data sets provided by NASA. The analytical analysis of the ETHERNET LAN and the video terminals (VTs) distribution system are presented. An interface analysis of the smart terminal network model which allows the data flow requirements due to VTs on the ETHERNET LAN to be estimated, is presented.

Ingels, F. M.↗

Cardea: Providing Support for Dynamic Resource Access in a Distributed Computing Environment

The environment framing the modem authorization process span domains of administration, relies on many different authentication sources, and manages complex attributes as part of the authorization process. Cardea facilitates dynamic access control within this environment as a central function of an inter-operable authorization framework. The system departs from the traditional authorization model by separating the authentication and authorization processes, distributing the responsibility for authorization data and allowing collaborating domains to retain control over their implementation mechanisms. Critical features of the system architecture and its handling of the authorization process differentiate the system from existing authorization components by addressing common needs not adequately addressed by existing systems. Continuing system research seeks to enhance the implementation of the current authorization model employed in Cardea, increase the robustness of current features, further the framework for establishing trust and promote interoperability with existing security mechanisms.

Lepro, Rebekah↗

Distributed computer system enhances productivity for SRB joint optimization

Initial calculations of a redesign of the solid rocket booster joint that failed during the shuttle tragedy showed that the design had a weight penalty associated with it. Optimization techniques were to be applied to determine if there was any way to reduce the weight while keeping the joint opening closed and limiting the stresses. To allow engineers to examine as many alternatives as possible, a system was developed consisting of existing software that coupled structural analysis with optimization which would execute on a network of computer workstations. To increase turnaround, this system took advantage of the parallelism offered by the finite difference technique of computing gradients to allow several workstations to contribute to the solution of the problem simultaneously. The resulting system reduced the amount of time to complete one optimization cycle from two hours to one-half hour with a potential of reducing it to 15 minutes. The current distributed system, which contains numerous extensions, requires one hour turnaround per optimization cycle. This would take four hours for the sequential system.

Rogers, James L., Jr.↗

Secure Large-Scale Airport Simulations Using Distributed Computational Resources

To fully conduct research that will support the far-term concepts, technologies and methods required to improve the safety of Air Transportation a simulation environment of the requisite degree of fidelity must first be in place. The Virtual National Airspace Simulation (VNAS) will provide the underlying infrastructure necessary for such a simulation system. Aerospace-specific knowledge management services such as intelligent data-integration middleware will support the management of information associated with this complex and critically important operational environment. This simulation environment, in conjunction with a distributed network of supercomputers, and high-speed network connections to aircraft, and to Federal Aviation Administration (FAA), airline and other data-sources will provide the capability to continuously monitor and measure operational performance against expected performance. The VNAS will also provide the tools to use this performance baseline to obtain a perspective of what is happening today and of the potential impact of proposed changes before they are introduced into the system.

McDermott, William J.↗

Deep Generative Models that Solve PDEs: Distributed Computing for Training Large Data-Free Models

Recent progress in scientific machine learning (SciML) has opened up the possibility of training novel neural network architectures that solve complex partial differential equations (PDEs). Several (nearly data free) approaches have been recently reported that successfully solve PDEs, with examples including deep feed forward networks, generative networks, and deep encoder-decoder networks. However, practical adoption of these approaches is limited by the difficulty in training these models, especially to make predictions at large output resolutions (≥1024×1024). Here we report on a software framework for data parallel distributed deep learning that resolves the twin challenges of training these large SciML models - training in reasonable time as well as distributing the storage requirements. Our framework provides several out of the box functionality including (a) loss integrity independent of number of processes, (b) synchronized batch normalization, and (c) distributed higher-order optimization methods. We show excellent scalability of this framework on both cloud as well as HPC clusters, and report on the interplay between bandwidth, network topology and bare metal vs cloud. We deploy this approach to train generative models of sizes hitherto not possible, showing that neural PDE solvers can be viably trained for practical applications. We also demonstrate that distributed higher-order optimization methods are 2-3× faster than stochastic gradient-based methods and provide minimal convergence drift with higher batch-size.

PDEs↗

Issues and recommendations associated with distributed computation and data management systems for the space sciences

The primary purpose of the report is to explore management approaches and technology developments for computation and data management systems designed to meet future needs in the space sciences.The report builds on work presented in previous reports on solar-terrestrial and planetary reports, broadening the outlook to all of the space sciences, and considering policy issues aspects related to coordiantion between data centers, missions, and ongoing research activities, because it is perceived that the rapid growth of data and the wide geographic distribution of relevant facilities will present especially troublesome problems for data archiving, distribution, and analysis.

Source record↗

Towards an Introspective Dynamic Model of Globally Distributed Computing Infrastructures

Large-scale scientific collaborations like ATLAS, Belle II, CMS, DUNE, and others involve hundreds of research institutes and thousands of researchers spread across the globe. These experiments generate petabytes of data, with volumes soon expected to reach exabytes. Consequently, there is a growing need for computation, including structured data processing from raw data to consumer-ready derived data, extensive Monte Carlo simulation campaigns, and a wide range of end-user analysis. To manage these computational and storage demands, centralized workflow and data management systems are implemented. However, decisions regarding data placement and payload allocation are often made disjointly and via heuristic means. A significant obstacle in adopting more effective heuristic or AI-driven solutions is the absence of a quick and reliable introspective dynamic model to evaluate and refine alternative approaches. In this study, we aim to develop such an interactive system using real-world data. By examining job execution records from the PanDA workflow management system, we have pinpointed key performance indicators such as queuing time, error rate, and the extent of remote data access. The dataset includes five months of activity. Additionally, we are creating a generative AI model to simulate time series of payloads, which incorporate visible features like category, event count, and submitting group, as well as hidden features like the total computational load—derived from existing PanDA records and computing site capabilities. These hidden features, which are not visible to job allocators, whether heuristic or AI-driven, influence factors such as queuing times and data movement.

kilic, Ozgur Ozan [Brookhaven National Laboratory ↗

Extending the distributed computing infrastructure of the CMS experiment with HPC resources

Particle accelerators are an important tool to study the fundamental properties of elementary particles. Currently the highest energy accelerator is the LHC at CERN, in Geneva, Switzerland. Each of its four major detectors, such as the CMS detector, produces dozens of Petabytes of data per year to be analyzed by a large international collaboration. The processing is carried out on the Worldwide LHC Computing Grid, that spans over more than 170 compute centers around the world and is used by a number of particle physics experiments. Recently the LHC experiments were encouraged to make increasing use of HPC resources. While Grid resources are homogeneous with respect to the used Grid middleware, HPC installations can be very different in their setup. In order to integrate HPC resources into the highly automatized processing setups of the CMS experiment a number of challenges need to be addressed. For processing, access to primary data and metadata as well as access to the software is required. At Grid sites all this is achieved via a number of services that are provided by each center. However at HPC sites many of these capabilities cannot be easily provided and have to be enabled in the user space or enabled by other means. At HPC centers there are often restrictions regarding network access to remote services, which is again a severe limitation. The paper discusses a number of solutions and recent experiences by the CMS experiment to include HPC resources in processing campaigns.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

CMS distributed computing workflow experience

The vast majority of the CMS Computing capacity, which is organized in a tiered hierarchy, is located away from CERN. The 7 Tier-1 sites archive the LHC proton-proton collision data that is initially processed at CERN. These sites provide access to all recorded and simulated data for the Tier-2 sites, via wide-area network (WAN) transfers. All central data processing workflows are executed at the Tier-1 level, which contain re-reconstruction and skimming workflows of collision data as well as reprocessing of simulated data to adapt to changing detector conditions. This paper describes the operation of the CMS processing infrastructure at the Tier-1 level. The Tier-1 workflows are described in detail. The operational optimization of resource usage is described. In particular, the variation of different workflows during the data taking period of 2010, their efficiencies and latencies as well as their impact on the delivery of physics results is discussed and lessons are drawn from this experience. The simulation of proton-proton collisions for the CMS experiment is primarily carried out at the second tier of the CMS computing infrastructure. Half of the Tier-2 sites of CMS are reserved for central Monte Carlo (MC) production while the other half is available for user analysis. This paper summarizes the large throughput of the MC production operation during the data taking period of 2010 and discusses the latencies and efficiencies of the various types of MC production workflows. We present the operational procedures to optimize the usage of available resources and we the operational model of CMS for including opportunistic resources, such as the larger Tier-3 sites, into the central production operation.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗