Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “CODE SHARING”

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 991 records · Page 55

BSE versus StarTrack: Implementations of new wind, remnant-formation, and natal-kick schemes in NBODY7 and their astrophysical consequences

As a result of their formation via massive single and binary stellar evolution, the masses of stellar-remnant black holes (BH) are subjects of great interest in this era of gravitational-wave detection from binary black hole (BBH) and binary neutron star merger events. In this work, we present new developments in the stellar-remnant formation and related schemes of the current N-body evolution program NBODY7. We demonstrate that the newly implemented stellar-wind and remnant-formation schemes in the stellar-evolutionary sector or BSE of the NBODY7 code, such as the “rapid” and the “delayed” supernova (SN) schemes along with an implementation of pulsational-pair-instability and pair-instability supernova (PPSN/PSN), now produce neutron star (NS) and BH masses that agree nearly perfectly, over large ranges of zero-age-main-sequence (ZAMS) mass and metallicity, with those from the widely recognised StarTrack population-synthesis program. We also demonstrate the new, recipe-based implementations of various widely debated mechanisms of natal kicks on NSs and BHs, such as “convection-asymmetry-driven”, “collapse-asymmetry-driven”, and “neutrino-emission-driven” kicks, in addition to a fully consistent implementation of the standard, fallback-dependent, momentum-conserving natal kick. All the above newly implemented schemes are also shared with the standalone versions of SSE and BSE. Furthermore, all these demonstrations are performed with both the updated standalone BSE and the updated NBODY7/BSE. When convolved with stellar and primordial-binary populations as observed in young massive clusters, such remnant-formation and natal-kick mechanisms crucially determine the accumulated number, mass, and mass distribution of the BHs retained in young massive, open, and globular clusters (GCs); these BHs would eventually become available for long-term dynamical processing. Among other conclusions, we find that although the newer, delayed SN remnant formation model gives birth to the largest number (mass) of BHs, the older remnant-formation schemes cause the largest number (mass) of BHs to survive in clusters, when incorporating SN material fallback onto the BHs. The SN material fallback also causes the convection-asymmetry-driven SN kick to effectively retain similar numbers and masses of BHs in clusters as for the standard, momentum-conserving kick. The collapse-asymmetry-driven SN kick would cause nearly all BHs to be retained in clusters irrespective of their mass, remnant-formation model, and metallicity, whereas the inference of a large population of BHs in GCs would potentially rule out the neutrino-driven SN kick mechanism. Pre-SN mergers of massive primordial binaries would potentially cause BH masses to deviate from the theoretical, single-star ZAMS to mass-remnant mass relation unless a substantial of the total merging stellar mass of up to ≈40% is lost during a merger process. In particular, such mergers, at low metallicities, have the potential to produce low-spinning BHs within the PSN mass gap that can be retained in a stellar cluster and be available for subsequent dynamical interactions. As recent studies indicate, the new remnant-formation modelling reassures us that young massive and open clusters would potentially contribute to the dynamical BBH merger detection rate to a similar extent as their more massive GC counterparts.

79 ASTRONOMY AND ASTROPHYSICS↗

Plasma Science and Innovation Center (PSI-Center) at Washington, Wisconsin, and Utah State, ARRA Supplement

The objective of the Plasma Science and Innovation Center (PSI-Center) is to develop and deploy computational models that simulate conditions in smaller, concept-exploration plasma experiments. The PSIC group at the University of Wisconsin-Madison, led by Prof. Carl Sovinec, uses and enhances the Non-Ideal Magnetohydrodynamics with Rotation, Open Discussion (NIMROD) code, to simulate macroscopic plasma dynamics in a number of magnetic confinement configurations. These numerical simulations provide information on how magnetic fields and plasma flows evolve over all three spatial dimensions, which supplements the limited access of diagnostics in plasma experiments. The information gained from simulation helps explain how plasma evolves. It is also used to engineer more effective plasma confinement systems, reducing the need for building many experiments to cover the physical parameter space. The ultimate benefit is a more cost-effective approach to the development of fusion energy for peaceful power production. The supplemental funds provided by the American Recovery and Reinvestment Act of 2009 were used to purchase computer components that were assembled into a 48-core system with 256 Gb of shared memory. The system was engineered and constructed by the group's system administrator at the time, Anthony Hammond. It was successfully used by then graduate student, Dr. John O'Bryan, for computing magnetic relaxation dynamics that occur during experimental tests of non-inductive startup in the Pegasus Toroidal Experiment (pegasus.ep.wisc.edu). Dr. O'Bryan's simulations provided the first detailed explanation of how the driven helical filament of electrical current evolves into a toroidal tokamak-like plasma configuration.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Automation of Data Traffic Control on DSM Architecture

The design of distributed shared memory (DSM) computers liberates users from the duty to distribute data across processors and allows for the incremental development of parallel programs using, for example, OpenMP or Java threads. DSM architecture greatly simplifies the development of parallel programs having good performance on a few processors. However, to achieve a good program scalability on DSM computers requires that the user understand data flow in the application and use various techniques to avoid data traffic congestions. In this paper we discuss a number of such techniques, including data blocking, data placement, data transposition and page size control and evaluate their efficiency on the NAS (NASA Advanced Supercomputing) Parallel Benchmarks. We also present a tool which automates the detection of constructs causing data congestions in Fortran array oriented codes and advises the user on code transformations for improving data traffic in the application.

Frumkin, Michael↗

Determination of Critical Experiment Correlations Using the Sampler Sequence Within SCALE 6.2

The validation of neutron transport methods used in nuclear criticality safety analyses is required by consensus American National Standards Institute/American Nuclear Society (ANSI/ANS) standards. In the last decade, there has been an increased interest in correlations among critical experiments used in validation that have shared physical attributes and which impact the independence of each measurement. The statistical methods included in many of the frequently cited guidance documents on performing validation calculations incorporate the assumption that all individual measurements are independent, so little guidance is available to practitioners on the topic. Typical guidance includes recommendations to select experiments from multiple facilities and experiment series in an attempt to minimize the impact of correlations or common-cause errors in experiments. Recent efforts have been made both to determine the magnitude of such correlations between experiments and to develop and apply methods for adjusting the bias and bias uncertainty to account for the correlations. This paper describes recent work performed at Oak Ridge National Laboratory using the Sampler sequence from the SCALE code system to develop experimental correlations using a Monte Carlo sampling technique. Sampler will be available for the first time with the release of SCALE 6.2, and a brief introduction to the methods used to calculate experiment correlations within this new sequence is presented in this paper. Techniques to utilize these correlations in the establishment of upper subcritical limits are the subject of a companion paper and will not be discussed here. Example experimental uncertainties and correlation coefficients are presented for a variety of low-enriched uranium water-moderated lattice experiments selected for use in a benchmark exercise by the Working Party on Nuclear Criticality Safety Subgroup on Uncertainty Analysis in Criticality Safety Analyses. The results include studies on the effect of fuel rod pitch on the correlations, and some observations are also made regarding difficulties in determining experimental correlations using the Monte Carlo sampling technique.

42 ENGINEERING↗

Determination of critical experiment correlations using the sampler sequence within scale 6.2 - 14602

The validation of neutron transport methods used in nuclear criticality safety analyses is required by consensus American National Standards Institute/American Nuclear Society (ANSI/ANS) standards. In the last decade, there has been an increased interest in correlations among critical experiments used in validation that have shared physical attributes and which impact the independence of each measurement. The statistical methods included in many of the frequently cited guidance documents on performing validation calculations incorporate the assumption that all individual measurements are independent, so little guidance is available to practitioners on the topic. Typical guidance includes recommendations to select experiments from multiple facilities and experiment series in an attempt to minimize the impact of correlations or common-cause errors in experiments. Recent efforts have been made both to determine the magnitude of such correlations between experiments and to develop and apply methods for adjusting the bias and bias uncertainty to account for the correlations. This paper describes recent work performed at Oak Ridge National Laboratory using the Sampler sequence from the SCALE code system to develop experimental correlations using a Monte Carlo sampling technique. Sampler will be available for the first time with the release of SCALE 6.2, and a brief introduction to the methods used to calculate experiment correlations within this new sequence is presented in this paper. Techniques to utilize these correlations in the establishment of upper subcritical limits are the subject of a companion paper and will not be discussed here. Example experimental uncertainties and correlation coefficients are presented for a variety of low-enriched uranium water-moderated lattice experiments selected for use in a benchmark exercise by the Working Party on Nuclear Criticality Safety Subgroup on Uncertainty Analysis in Criticality Safety Analyses. The results include studies on the effect of fuel rod pitch on the correlations, and some observations are also made regarding difficulties in determining experimental correlations using the Monte Carlo sampling technique. (authors)

42 ENGINEERING↗

Epi Archive: Automated Synthesis of Global Notifiable Disease Data

ObjectiveLANL has built software that automatically collects global notifiable disease data, synthesizes the data, and makes it available to humans and computers within the Biosurveillance Ecosystem (BSVE) as a novel data stream. These data have many applications including improving the prediction and early warning of disease events.IntroductionMost countries do not report national notifiable disease data in a machine-readable format. Data are often in the form of a file that contains text, tables and graphs summarizing weekly or monthly disease counts. This presents a problem when information is needed for more data intensive approaches to epidemiology, biosurveillance and public health.While most nations likely store incident data in a machine-readable format, governments are often hesitant to share data openly for a variety of reasons that include technical, political, economic, and motivational issues1.A survey conducted by LANL of notifiable disease data reporting in over fifty countries identified only a few websites that report data in a machine-readable format. The majority (>70%) produce reports as PDF files on a regular basis. The bulk of the PDF reports present data in a structured tabular format, while some report in natural language.The structure and format of PDF reports change often; this adds to the complexity of identifying and parsing the desired data. Not all websites publish in English, and it is common to find typos and clerical errors.LANL has developed a tool, Epi Archive, to collect global notifiable disease data automatically and continuously and make it uniform and readily accessible.MethodsWe conducted a survey of the national notifiable disease reporting systems notating how the data are reported and in what formats. We determined the minimal metadata that is required to contextualize incident counts properly, as well as optional metadata that is commonly found.The development of software to regularly ingest notifiable disease data and make it available involves three or four main steps: scraping, detecting, parsing and persisting.Scraping: we examine website design and determine reporting mechanisms for each country/website, as well as what varies across the reporting mechanisms. We then designed and wrote code to automate the downloading of the data for each country. We store all artifacts presented as files (PDF, XLSX, etc.) in their original form, along with appropriate metadata for parsing and data provenance.Detecting: This step is required when parsing structured non-machine-readable data such as tabular data in PDF files. We combined the Nurminen methodology of PDF table detection with in-house heuristics to find the desired data within PDF reports2.Parsing: We determined what to extract from each dataset and parsed these data into uniform data structures, correctly accommodating the variations in metadata (e.g., time interval definitions) and the various human languages.Persisting: We store the data in the Epi Archive database and make it available on the internet and through the BSVE. The data is persisted into a structured and normalized SQL database.ResultsThe Epi Archive tool currently contains national and/or subnational notifiable disease data from twenty nations. When a user accesses the Epi Archive site, they are prompted with four fields: country, subregion, disease of interest, and date duration. Upon form submission, a time series is generated from the users’ specifications. The generated graph can then be downloaded into a CSV file if a user is interested in performing personal analysis. Additionally, the data from Epi Archive can be reached through a REST API (Representational State Transfer Application Programming Interface).ConclusionsLANL, as part of a currently funded DTRA effort, is automatically and continually collecting global notifiable disease data. While 20 nations are in production, more are being brought online in the near future. These data are already being utilized and will have many applications including improving the prediction and early warning of disease events.References[1] van Panhuis WG, Paul P, Emerson C, et al. A systematic review of barriers to data sharing in public health. BMC Public Health. 2014. 14:1144. doi:10.1186/1471-2458-14-1144[2] Nurminen, Anssi. "Algorithmic extraction of data in tables in PDF documents." (2013).

Khalsa, Hari S.↗

Leveraging The Open Provenance Model as a Multi-Tier Model for Global Climate Research

Global climate researchers rely upon many forms of sensor data and analytical methods to help profile subtle changes in climate conditions. The U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) program provides researchers with curated Value Added Products (VAPs) resulting from continuous sensor data streams, data fusion, and modeling. The ARM operations staff and software development teams (data producers) rely upon a number of techniques to ensure strict quality control (QC) and quality assurance (QA) standards are maintained. Climate researchers (data consumers) are highly interested in obtaining as much provenance (data quality, data pedigree) as possible to establish data trustworthiness. Currently all the provenance is not easily attainable or identifiable without significant efforts to extract and piece together information from configuration files, log files, codes, and status information from ARM databases. The need for a formalized approach to managing provenance became paramount with the planned addition of 120 new instruments, new data products, and data collection scaling to half a terabyte daily. Last year our research identified the need for a multi-tier provenance model to enable the data consumer easy access to the provenance for their data. This year we are leveraging the Open Provenance Model as a foundational construct that serves the needs of both the VAP producers and consumers, we are organizing the provenance in different tiers of granularity to model VAP lineage, causality at the component level within a VAP, and the causality for each time step as samples are being assembled within the VAP. This paper shares our implementation strategy and how the ARM operations staff and the climate research community can greatly benefit from this approach to more effectively assess and quantify VAP provenance.

54 ENVIRONMENTAL SCIENCES↗

Fenestration System Performance Research, Testing, and Evaluation

The US DOE was and is instrumental to NFRC's beginning and its continued success. The 2005 to 2009 funding enables NFRC to continue expanding and create new, improved ratings procedures. Research funded by the US DOE enables increased fenestration energy rating accuracy. International harmonization efforts supported by the US DOE allow the US to be the global leader in fenestration energy ratings. Many other governments are working with the NFRC to share its experience and knowledge toward development of their own national fenestration rating process similar to the NFRC's. The broad and diverse membership composition of NFRC allows anyone with a fenestration interest to come forward with an idea or improvement to the entire fenestration community for consideration. The NFRC looks forward to the next several years of growth while remaining the nation's resource for fair, accurate, and credible fenestration product energy ratings. NFRC continues to improve its rating system by considering new research, methodologies, and expanding to include new fenestration products. Currently, NFRC is working towards attachment energy ratings. Attachments are blinds, shades, awnings, and overhangs. Attachments may enable a building to achieve significant energy savings. An NFRC rating will enable fair competition, a basis for code references, and a new ENERGY STAR product category. NFRC also is developing rating methods to consider non specular glazing such as fritted glass. Commercial applications frequently use fritted glazing, but no rating method exists. NFRC is testing new software that may enable this new rating and contribute further to energy conservation. Around the world, many nations are seeking new energy conservation methods and NFRC is poised to harmonize its rating system assisting these nations to better manage and conserve energy in buildings by using NFRC rated and labeled fenestration products. As this report has shown, much more work needs to be done to continues research to improve existing ratings and develop new ones. NFRC needs to continue the work it has begun in several nations to implement the NFRC rating system that has been introduced. Many nations are eager to accept the expertise NFRC can offer to achieve energy conservation goals. NFRC looks forward to a continues partnership with the US Department of Energy to cooperatively achieve both.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

UPC++ as_eager Working Group Draft, Revision 2020.6.2

This draft proposes an extension for a new future-based completion variant that can be more effectively streamlined for RMA and atomic access operations that happen to be satisfied at runtime using purely node-local resources. Many such operations are most efficiently performed synchronously using load/store instructions on shared-memory mappings, where the actual access may only require a few CPU instructions. In such cases we believe it’s critical to minimize the overheads imposed by the UPC++ runtime and completion queues, in order to enable efficient operation on hierarchical node hardware using shared-memory bypass. The new upcxx::{source,operation}_cx::as_eager_future() completion variant accomplishes this goal by relaxing the current restriction that future-returning access operations must return a non-ready future whose completion is deferred until a subsequent explicit invocation of user-level progress. This relaxation allows access operations that are completed synchronously to instead return a ready future, thereby avoiding most or all of the runtime costs associated with deferment of future completion and subsequent mandatory entry into the progress engine. We additionally propose to make this new as_eager_future() completion variant the new default completion for communication operations that currently default to returning a future. This should encourage use of the streamlined variant, and may provide performance improvements to some codes without source changes. A mechanism is proposed to restore the legacy behavior on-demand for codes that might happen to rely on deferred completion for correctness. Finally, we propose a new as_eager_promise() completion variant that extends analogous improvements to promise-based completion, and corresponding changes to the default behavior of as_promise().

97 MATHEMATICS AND COMPUTING↗

Massive hybrid parallelism for fully implicit multiphysics

As hardware advances continue to modify the supercomputing landscape, traditional scientific software development practices will become more outdated, ineffective, and inefficient. The process of rewriting/retooling existing software for new architectures is a Sisyphean task, and results in substantial hours of development time, effort, and money. Software libraries which provide an abstraction of the resources provided by such architectures are therefore essential if the computational engineering and science communities are to continue to flourish in this modern computing environment. The Multiphysics Object Oriented Simulation Environment (MOOSE) framework enables complex multiphysics analysis tools to be built rapidly by scientists, engineers, and domain specialists, while also allowing them to both take advantage of current HPC architectures, and efficiently prepare for future supercomputer designs. MOOSE employs a hybrid shared-memory and distributed-memory parallel model and provides a complete and consistent interface for creating multiphysics analysis tools. In this paper, a brief discussion of the mathematical algorithms underlying the framework and the internal object-oriented hybrid parallel design are given. Representative massively parallel results from several applications areas are presented, and a brief discussion of future areas of research for the framework are provided. (authors)

97 MATHEMATICS AND COMPUTING↗

Use of the portable decay gamma source feature in the R2Smesh code system for the accurate estimation of photon radiation dose fields around activated ITER components - Paper 111

The R2Smesh ('Rigorous 2-step') code system, developed by KIT for the determination of shut-down dose rate distribution on mesh grid, can be utilized to provide photon radiation dose field around activated components which are moved from the irradiation site in fusion reactor to another external location. R2Smesh couples MCNP transport calculations and FISPACT inventory calculations through suitable interfaces utilizing MCNP mesh tally capability. Neutron flux and the decay gamma source are provided with high resolution on the mesh grid superimposed to the real geometry to take into account the spatial variations of the flux and the decay gamma source distributions. This feature enables the decay gamma source distribution to be transferred from the irradiation site in the reactor to any other external location for the determination of photon flux and dose rate distributions. In this work, we applied the feature of the portable decay gamma source in R2Smesh code system to a dose rate analysis for a tunnel connecting the ITER tokamak complex building (TCB) and the hot cell facility building (HCFB). This paper presents the use of the portability feature of the decay gamma source in R2Smesh interface code system developed by KIT for a dose rate analysis of the ITER tunnel connecting TCB and HCFB. This application shows that the decay gamma source of the activated components could be transferred successfully from one coordinate system to another coordinate system and superimposed to the geometry of the components. This is very useful when the activated components need to be transported from the reactor to somewhere else for maintenance, testing or storage and the dose rate analysis is requested. Another portable feature, which could convert the 3D decay gamma source distribution generated by R2Smesh system to the MCNP 'SDEF' card source description and avoid the use of source sampling subroutine, is under consideration. With this feature, one decay gamma source result for the specific reactor components could be shared by different institutions who need to use the same decay gamma source for different analysis. (authors)

61 RADIATION PROTECTION AND DOSIMETRY↗

Enabling end-to-end secure federated learning in biomedical research on heterogeneous computing environments with APPFLx

Facilitating large-scale, cross-institutional collaboration in biomedical machine learning (ML) projects requires a trustworthy and resilient federated learning (FL) environment to ensure that sensitive information such as protected health information is kept confidential. Specifically designed for this purpose, this work introduces APPFLx - a low-code, easy-to-use FL framework that enables easy setup, configuration, and running of FL experiments. APPFLx removes administrative boundaries of research organizations and healthcare systems while providing secure end-to-end communication, privacy-preserving functionality, and identity management. Furthermore, it is completely agnostic to the underlying computational infrastructure of participating clients, allowing an instantaneous deployment of this framework into existing computing infrastructures. Experimentally, the utility of APPFLx is demonstrated in two case studies: (1) predicting participant age from electrocardiogram (ECG) waveforms, and (2) detecting COVID-19 disease from chest radiographs. Here, ML models were securely trained across heterogeneous computing resources, including a combination of on-premise high-performance computing and cloud computing facilities. By securely unlocking data from multiple sources for training without directly sharing it, these FL models enhance generalizability and performance compared to centralized training models while ensuring data remains protected. In conclusion, APPFLx demonstrated itself as an easy-to-use framework for accelerating biomedical studies across organizations and healthcare systems on large datasets while maintaining the protection of private medical data.

Biomedical Research↗

Optimal Bayesian supervised domain adaptation for RNA sequencing data

Abstract Motivation When learning to subtype complex disease based on next-generation sequencing data, the amount of available data is often limited. Recent works have tried to leverage data from other domains to design better predictors in the target domain of interest with varying degrees of success. But they are either limited to the cases requiring the outcome label correspondence across domains or cannot leverage the label information at all. Moreover, the existing methods cannot usually benefit from other information available a priori such as gene interaction networks. Results In this article, we develop a generative optimal Bayesian supervised domain adaptation (OBSDA) model that can integrate RNA sequencing (RNA-Seq) data from different domains along with their labels for improving prediction accuracy in the target domain. Our model can be applied in cases where different domains share the same labels or have different ones. OBSDA is based on a hierarchical Bayesian negative binomial model with parameter factorization, for which the optimal predictor can be derived by marginalization of likelihood over the posterior of the parameters. We first provide an efficient Gibbs sampler for parameter inference in OBSDA. Then, we leverage the gene-gene network prior information and construct an informed and flexible variational family to infer the posterior distributions of model parameters. Comprehensive experiments on real-world RNA-Seq data demonstrate the superior performance of OBSDA, in terms of accuracy in identifying cancer subtypes by utilizing data from different domains. Moreover, we show that by taking advantage of the prior network information we can further improve the performance. Availability and implementation The source code for implementations of OBSDA and SI-OBSDA are available at the following link. https://github.com/SHBLK/BSDA. Supplementary information Supplementary data are available at Bioinformatics online.

Biochemistry & Molecular Biology↗

What Employees Need (and Want) to Hear When Justifying the Suspension of a Regulated Metals Plan for the Processing of Drums Containing Metal Turnings

A Regulated Metals Plan (RMP) was implemented for outdoor work activities involving the removal and disposition of approximately 4,000 deteriorated waste drums containing 236 metric tonnes (260 tons) of lead turnings from various, unspecified machine shop facilities at the Paducah Gaseous Diffusion Plant. Until exposure monitoring could prove otherwise, the work area established for processing the drums was conservatively defined as a Lead Regulated Area (LRA) subject to the Occupational Safety and Health Administration's Lead Standard found in Title 29 of the Code of Federal Regulations, Part 1910.1025. The vast majority of the analytical results for the industrial hygiene breathing zone samples collected and tested for arsenic, beryllium, cadmium, chromium, lead, nickel, selenium, silver, and thallium using the National Institute for Occupational Safety and Health's analytical method 7300 were equivalent to the laboratory detection limits for each analyte. All results were less than 6% of their respective Permissible Exposure Limits (PEL), except for one nickel result that was approximately 17% of its PEL. The results provided justification to eventually down-post the LRA to existing employee protection requirements. In addition to removing the deteriorated drums and accompanying debris, the success of this project was quantified in terms of zero recordable injuries. The primary contributor in achieving this success was the sharing and communication of information between management, safety, and the field teams. Specifically, this was what the employees needed (and wanted) to hear when justifying the suspension of the RMP for the processing of drums containing metal turnings. Daily briefings on the status of the project and field monitoring results were just as important as maintaining budget and schedule milestones. Also, the Environmental, Safety and Health organization maintained its presence by continuing to monitor evolving field conditions to ensure the effectiveness of its plans and procedures. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Reduced-model scrape-off layer turbulence (nSOLT) simulations comparing three fueling scenarios

The 2D scrape-off-layer turbulence code (nSOLT) includes 1D Boltzmann neutral–plasma interactions, a model of divertor recycling (introduced here), and a fixed source of plasma concentrated at the core-side boundary. Three fueling methods are considered herein: (1) neutral injection in the far-SOL is accomplished by specifying the density of Franck–Condon distributed neutrals streaming in from the boundary. (2) Divertor recycling is modeled by injecting a fraction of the particle parallel flux in the scrape-off layer (SOL) back into the edge region as a source of plasma. (3) A constant source fuels the edge plasma from the core-side boundary to model pellet injection. For machine parameters (B, R m , and L // ) illustrative of the MAST-U device, and for a deuterium plasma, turbulent equilibria are obtained that share the same plasma fueling rate for each of the three fueling methods, with only one of the sources on in each case. In the presence of self-consistent turbulence, quasi-steady plasma and neutral (deuterium) profiles, fueling efficiencies, SOL transparencies, and heat flux widths are compared. Characteristics of the turbulent fluctuations, including skewness, cross-phases, and power spectra, are described. The calculated fueling efficiencies, SOL transparencies to neutral penetration, and many of the turbulent properties are remarkably similar for all three fueling methods despite significant differences in the plasma profiles. The nonlinear states of the three cases are dominated by separatrix-spanning vortex cells that control particle and heat losses into the SOL.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Integrating Cache Performance Modeling and Tuning Support in Parallelization Tools

With the resurgence of distributed shared memory (DSM) systems based on cache-coherent Non Uniform Memory Access (ccNUMA) architectures and increasing disparity between memory and processors speeds, data locality overheads are becoming the greatest bottlenecks in the way of realizing potential high performance of these systems. While parallelization tools and compilers facilitate the users in porting their sequential applications to a DSM system, a lot of time and effort is needed to tune the memory performance of these applications to achieve reasonable speedup. In this paper, we show that integrating cache performance modeling and tuning support within a parallelization environment can alleviate this problem. The Cache Performance Modeling and Prediction Tool (CPMP), employs trace-driven simulation techniques without the overhead of generating and managing detailed address traces. CPMP predicts the cache performance impact of source code level "what-if" modifications in a program to assist a user in the tuning process. CPMP is built on top of a customized version of the Computer Aided Parallelization Tools (CAPTools) environment. Finally, we demonstrate how CPMP can be applied to tune a real Computational Fluid Dynamics (CFD) application.

Waheed, Abdul↗

Portals 4 network API definition and performance measurement

Portals is a low-level network programming interface for distributed memory massively parallel computing systems designed by Sandia, UNM, and Intel. Portals has been designed to provide high message rates and to provide the flexibility to support a variety of higher-level communication paradigms. This project developed and analyzed an implementation of Portals using shared memory in order to measure and understand the impact of using general-purpose compute cores to handle network protocol processing functions. The goal of this study was to evaluate an approach to high-performance networking software design and hardware support that would enable important DOE modeling and simulation applications to perform well and to provide valuable input to Intel so they can make informed decisions about future network software and hardware products that impact DOE applications.

99 GENERAL AND MISCELLANEOUS↗