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

Title: Taking Lessons Learned from a Proxy Application to a Full Application for SNAP and PARTISN

Abstract

SNAP is a proxy application which simulates the computational motion of a neutral particle transport code, PARTISN. Here in this work, we have adapted parts of SNAP separately; we have re-implemented the iterative shell of SNAP in the task-model runtime Legion, showing an improvement to the original schedule, and we have created multiple Kokkos implementations of the computational kernel of SNAP, displaying similar performance to the native Fortran. We then translate our Kokkos experiments in SNAP to PARTISN, necessitating engineering development, regression testing, and further thought.

Authors:
ORCiD logo [1];  [1];  [1]
  1. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1418763
Report Number(s):
LA-UR-16-29442
Journal ID: ISSN 1877-0509
Grant/Contract Number:
AC52-06NA25396
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Procedia Computer Science
Additional Journal Information:
Journal Volume: 108; Conference: International Conference on Computational Science 2017, Zurich (Switzerland), 12-14 Jun 2017; Journal ID: ISSN 1877-0509
Publisher:
Elsevier
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING

Citation Formats

Womeldorff, Geoffrey Alan, Payne, Joshua Estes, and Bergen, Benjamin Karl. Taking Lessons Learned from a Proxy Application to a Full Application for SNAP and PARTISN. United States: N. p., 2017. Web. doi:10.1016/j.procs.2017.05.243.
Womeldorff, Geoffrey Alan, Payne, Joshua Estes, & Bergen, Benjamin Karl. Taking Lessons Learned from a Proxy Application to a Full Application for SNAP and PARTISN. United States. doi:10.1016/j.procs.2017.05.243.
Womeldorff, Geoffrey Alan, Payne, Joshua Estes, and Bergen, Benjamin Karl. Fri . "Taking Lessons Learned from a Proxy Application to a Full Application for SNAP and PARTISN". United States. doi:10.1016/j.procs.2017.05.243. https://www.osti.gov/servlets/purl/1418763.
@article{osti_1418763,
title = {Taking Lessons Learned from a Proxy Application to a Full Application for SNAP and PARTISN},
author = {Womeldorff, Geoffrey Alan and Payne, Joshua Estes and Bergen, Benjamin Karl},
abstractNote = {SNAP is a proxy application which simulates the computational motion of a neutral particle transport code, PARTISN. Here in this work, we have adapted parts of SNAP separately; we have re-implemented the iterative shell of SNAP in the task-model runtime Legion, showing an improvement to the original schedule, and we have created multiple Kokkos implementations of the computational kernel of SNAP, displaying similar performance to the native Fortran. We then translate our Kokkos experiments in SNAP to PARTISN, necessitating engineering development, regression testing, and further thought.},
doi = {10.1016/j.procs.2017.05.243},
journal = {Procedia Computer Science},
number = ,
volume = 108,
place = {United States},
year = {Fri Jun 09 00:00:00 EDT 2017},
month = {Fri Jun 09 00:00:00 EDT 2017}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:
  • Geothermal trade papers and the general press have called attention to the Baca Project - a joint project of the US Department of Energy (DOE), Union Geothermal Company of New Mexico, and Public Service of New Mexico - and particularly to the lack of sufficient steam that caused cancellation of the planned 50 MWe demonstration plant. A study of the difficulties that Union was encountering in developing the expected steam from its new production wells is presented. The Baca Reservoir Definition Review Team closely followed activities until the principal parties mutually agreed in January 1982 to terminate the project. Themore » review team issued its final report later that year. In contrast to the negative outcome of the project, the less newsworthy but beneficial scientific aspects of the project may have been largely ignored except by interested specialists in industry and research. Some of the lessons and information coming out of the project are discussed.« less
  • In the short time that plant-specific, full-scope probabilistic risk assessments (PRAs) have been performed, extensive progress has been made in understanding and managing risk. This paper discusses the various lessons learned during this time. Twenty-one specific PRAs are evaluated.
  • Probabilistic risk assessment (PRA) has been under development for over 20 yr, but it reached the level of widespread use only in the aftermath of the Three Mile Island accident. Over 30 PRAs have now been completed in the US. They have been in the mainstream of many licensing decisions because both the Nuclear Regulatory Commission and the nuclear utilities recognize that PRAs provide independent and comprehensive plant safety audits. Although some difficulties have been experienced, they have led to such efforts as interpretive or intercomparison studies. Numerous global and plant-specific insights have been derived. A new application termed riskmore » management is clearly emerging.« less
  • Many utilities are considering offering optional time-of-use (TOU) rates to residential customers. Because of the unique analytical problems involved in estimating customer response to voluntary TOU rates, the vast literature on the results from mandatory residential TOU experiments is of limited usefulness in guiding utilities in designing and marketing optional TOU rates. This paper reports on the voluntary residential TOU experiment recently concluded by Pacific Gas and Electric Company. The summary of experimental findings will prove useful to other utilities about to offer similar TOU rate options to residential customers.
  • The Sandia National Laboratories/New Mexico (SNL/NM) Environmental Restoration Project is halfway through excavating the Classified Waste Landfill in Technical Area II, a disposal area for weapon components for approximately 40 years. While the planning phase of any project is important, it is only a means of getting to the field implementation phase where reality quickly sinks in. Documents outlining the general processes are developed, heavy equipment, supply needs, requisite skills, and staffing levels are anticipated, and contingencies for waste management are put in place. However, the nature of landfill excavation dictates that even the most detailed plans will probably change.more » This project is proving that trying to account for undefined variables and predicting the total cost of landfill remediation is very difficult if the contents are not well known. In landfill excavation, contingency cannot be minimized. During development of the waste management plan, it was recognized that even the best forecasting could not formulate the perfect cradle-to-grave processes because waste streams are rarely definable before excavation begins. Typically, as excavation progresses and waste streams are generated, new characterization information allows further definition of disposal options which, in turn, modify the generation/management process. A general plan combined with close involvement of waste management personnel to resolve characterization and packaging questions during generation has worked very well. And, as expected, each new pit excavated creates new waste management challenges. The material excavated consists primarily of classified weapon assemblies and related components, so disposition must include demilitarization and sanitization. The demilitarization task at the start of the project was provided by an SNL/NM group that has since lost their funding and operational capability. This project is having to take on the task of disassembly, destruction, and recycling of classified components, along with the associated costs and infrastmcture. Very stringent radiological controls were imposed on site operations during the planning phase. Radiological controls that are not justified significantly impact the efficiency and cost of operations. If the initial approach is too conservative, there should be well-defined provisions for scaling down the protective measures to reflect the actual risks. Once the effectiveness of early detection, monitoring, and surveys is proven, radiological controls and postings should be re-evaluated to verify that they are appropriate. High levels of heavy metals dust were not anticipated during the planning phase but were suspected, then confirmed, during material handling. Respiratory protection and monitoring were upgraded accordingly and the costs added to the baseline. In contrast to radiological constraints, industrial hygiene guidelines were worked into the process with a minimum of adverse impact. While a lot of unforeseen expenses occur, some expected costs can be reduced. During the planning phase, the anticipated need to adequately characterize a variety of radionuclides in soil led to using Large Area Gamma Spectroscopy (LAGS) to survey all the soil excavated. About a quarter of the way through the project, it was obvious that very little radioactive material was present in the excavated soils. Since all the soil is processed through a screen plant, producing a fairly homogeneous mix, a more common method of sampling soil piles was implemented to replace the LAGS unit, increase productivity, and reduce costs. In summary, the most important lesson is to expect and be ready to change. Excavating a landfill requires the flexibility to quickly adjust processes to handle the unknown variables, and close attention to detail so all the different facets of the project are kept under control.« less