skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: FY06 LDRD Final Report Data Intensive Computing

Abstract

The goal of the data intensive LDRD was to investigate the fundamental research issues underlying the application of High Performance Computing (HPC) resources to the challenges of data intensive computing. We explored these issues through four targeted case studies derived from growing LLNL programs: high speed text processing, massive semantic graph analysis, streaming image feature extraction, and processing of streaming sensor data. The ultimate goal of this analysis was to provide scalable data management algorithms to support the development of a predictive knowledge capability consistent with the direction of Aurora.

Authors:
Publication Date:
Research Org.:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
902242
Report Number(s):
UCRL-TR-228102
TRN: US200717%%500
DOE Contract Number:
W-7405-ENG-48
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ALGORITHMS; LAWRENCE LIVERMORE NATIONAL LABORATORY; MANAGEMENT; PERFORMANCE; PROCESSING; VELOCITY

Citation Formats

Abdulla, G M. FY06 LDRD Final Report Data Intensive Computing. United States: N. p., 2007. Web. doi:10.2172/902242.
Abdulla, G M. FY06 LDRD Final Report Data Intensive Computing. United States. doi:10.2172/902242.
Abdulla, G M. Tue . "FY06 LDRD Final Report Data Intensive Computing". United States. doi:10.2172/902242. https://www.osti.gov/servlets/purl/902242.
@article{osti_902242,
title = {FY06 LDRD Final Report Data Intensive Computing},
author = {Abdulla, G M},
abstractNote = {The goal of the data intensive LDRD was to investigate the fundamental research issues underlying the application of High Performance Computing (HPC) resources to the challenges of data intensive computing. We explored these issues through four targeted case studies derived from growing LLNL programs: high speed text processing, massive semantic graph analysis, streaming image feature extraction, and processing of streaming sensor data. The ultimate goal of this analysis was to provide scalable data management algorithms to support the development of a predictive knowledge capability consistent with the direction of Aurora.},
doi = {10.2172/902242},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Tue Feb 13 00:00:00 EST 2007},
month = {Tue Feb 13 00:00:00 EST 2007}
}

Technical Report:

Save / Share:
  • This project was aimed at developing computational methods and tools to decipher the universal language of gene regulation encoded in genomes. It was proposed to identify and decipher the code and to functionally annotate genomic elements that orchestrate the temporal and spatial dynamics of gene expression in living cells.
  • We have taken the initiative to examine whether experiments on HED facilities, present and future, could achieve the extreme scaled conditions relevant to accreting neutron star atmospheres and accretion disks around black holes. The preliminary conclusion from this detailed scaling assessment is that if an exact scaled version of the photon bubble instability physics is desired, this will require experiments with (simultaneously) spatial scales of order {approx}1 mm, temperatures of order {approx}5 keV, magnetic fields of order a hundred megaGauss, and time scales of order several hundred psec. Aspects (subsets) of this physics can be studied under less demanding conditions.more » To achieve the temperatures required in targets of order several optical depths, we come to the preliminary conclusion that we would require an energy source that delivers of order of a megajoule of energy into a high Z target. A conceptual design for such an experiment could be to use the energy from a high gain ignition NIF capsule as our principle source of heating and acceleration whereby the target is in close proximity to the ignition capsule and then use external petawatt lasers to develop the magnetic fields required.« less
  • This is the final report for LDRD 01-ERD-005. The Principle Investigator was Robert Sharpe. Collaborators included Niel Madsen, Benjamin Fasenfest, John D. Rockway, of the Defense Sciences Engineering Division (DSED), Vikram Jandhyala and James Pingenot from the University of Washington, and Mark Stowell of the Center for Applications Development and Software Engineering (CADSE). It should be noted that Benjamin Fasenfest and Mark Stowell were partially supported under other funding. The purpose of this LDRD effort was to enhance LLNL's computational electromagnetics capability in the area of broadband radiation and scattering. For radiation and scattering problems our transient EM codes aremore » limited by the approximate Radiation Boundary Conditions (RBC's) used to model the radiation into an infinite space. Improved RBC's were researched, developed, and incorporated into the existing EMSolve finite-element code to provide a 10-100x improvement in the accuracy of the boundary conditions. Section I provides an introduction to the project and the project goals. Section II provides a summary of the project's research and accomplishments as presented in the attached papers.« less
  • The original goal of our research was to open up a new class of scientific experiments by increasing the power of newly available x-ray sources by orders of magnitude. This was accomplished by developing a new generation of x-ray optics, based on hard x-ray (10-200 keV) reflective and diffractive focusing elements. The optical systems we envision begin with a core reflective optic, which has the ability to capture and concentrate x-rays across a wide range of energies and angles band, combined with diffractive optics, based on large-scale multilayer structures, that will further enhance the spatial, spectral and temporal resolving powermore » of the system. Enabling technologies developed at LLNL such as precise mounting of thermally formed substrates, smoothing techniques and multilayer films of ultra-high reflectance and precision were crucial in the development and demonstration of our research objectives. Highlights of this phase of the project include: the design and fabrication of a concentrator optic for the Pleiades Thomson X-ray source located at LLNL, smoothing of glass substrates through application of polyimide films, and the design, fabrication and testing of novel volume multilayers structures. Part of our research into substrate smooth led to the development of a new technique (patent pending) to construct high-quality, inexpensive x-ray optics. This innovation resulted in LLNL constructing a x-ray optic for the CERN Axion Solar Telescope (CAST) and allowed LLNL to join the international experiment.« less
  • This project's work resulted in the following research projects: (1) BORG - Block-reORGanization for Self-optimizing Storage Systems; (2) ABLE - Active Block Layer Extensions; (3) EXCES - EXternal Caching in Energy-Saving Storage Systems; (4) GRIO - Guaranteed-Rate I/O Scheduler. These projects together help in substantially advancing the over-arching project goal of developing 'QoS-Enabled, High-Performance Storage Systems'.