Search NASA⌕ Search

SEARCH · Search NASA

Results for “concurrency programming”

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 325 records · Page 18

Wide Area Security Region Final Report

This report develops innovative and efficient methodologies and practical procedures to determine the wide-area security region of a power system, which take into consideration all types of system constraints including thermal, voltage, voltage stability, transient and potentially oscillatory stability limits in the system. The approach expands the idea of transmission system nomograms to a multidimensional case, involving multiple system limits and parameters such as transmission path constraints, zonal generation or load, etc., considered concurrently. The security region boundary is represented using its piecewise approximation with the help of linear inequalities (so called hyperplanes) in a multi-dimensional space, consisting of system parameters that are critical for security analyses. The goal of this approximation is to find a minimum set of hyperplanes that describe the boundary with a given accuracy. Methodologies are also developed to use the security hyperplanes, pre-calculated offline, to determine system security margins in real-time system operations, to identify weak elements in the system, and to calculate key contributing factors and sensitivities to determine the best system controls in real time and to assist in developing remedial actions and transmission system enhancements offline . A prototype program that automates the simulation procedures used to build the set of security hyperplanes has also been developed. The program makes it convenient to update the set of security hyperplanes necessitated by changes in system configurations. A prototype operational tool that uses the security hyperplanes to assess security margins and to calculate optimal control directions in real time has been built to demonstrate the project success. Numerical simulations have been conducted using the full-size Western Electricity Coordinating Council (WECC) system model, and they clearly demonstrated the feasibility and the effectiveness of the developed technology. Recommendations for the future work have also been formulated.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Feasibility of low-dose single-view 3D fiducial tracking concurrent with external beam delivery

Purpose: In external-beam radiation therapy, existing on-board x-ray imaging chains orthogonal to the delivery beam cannot recover 3D target trajectories from a single view in real-time. This limits their utility for real-time motion management concurrent with beam delivery. To address this limitation, the authors propose a novel concept for on-board imaging based on the inverse-geometry Scanning-Beam Digital X-ray (SBDX) system and evaluate its feasibility for single-view 3D intradelivery fiducial tracking. Methods: A chest phantom comprising a posterior wall, a central lung volume, and an anterior wall was constructed. Two fiducials were placed along the mediastinal ridge between the lung cavities: a 1.5 mm diameter steel sphere superiorly and a gold cylinder (2.6 mm length x 0.9 mm diameter) inferiorly. The phantom was placed on a linear motion stage that moved sinusoidally. Fiducial motion was along the source-detector (z) axis of the SBDX system with {+-}10 mm amplitude and a programmed period of either 3.5 s or 5 s. The SBDX system was operated at 15 frames per second, 100 kVp, providing good apparent conspicuity of the fiducials. With the stage moving, detector data were acquired and subsequently reconstructed into 15 planes with a 12 mm plane-to-plane spacing using digital tomosynthesis. A tracking algorithm was applied to the image planes for each temporal frame to determine the position of each fiducial in (x,y,z)-space versus time. A 3D time-sinusoidal motion model was fit to the measured 3D coordinates and root mean square (RMS) deviations about the fitted trajectory were calculated. Results: Tracked motion was sinusoidal and primarily along the source-detector (z) axis. The RMS deviation of the tracked z-coordinate ranged from 0.53 to 0.71 mm. The motion amplitude derived from the model fit agreed with the programmed amplitude to within 0.28 mm for the steel sphere and within -0.77 mm for the gold seed. The model fit periods agreed with the programmed periods to within 7%. Conclusions: Three dimensional fiducial tracking with approximately 1 mm or better accuracy and precision appears to be feasible with SBDX, supporting its use to guide radiotherapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Five-year summary and evaluation of operations and performance of the Utica aquifer and North Lake Basin Wetlands restoration project in 2004-2009.

This document reviews the performance of the groundwater (and wetlands) restoration program implemented by the Commodity Credit Corporation of the U.S. Department of Agriculture (CCC/USDA) at the former CCC/USDA grain storage facility in Utica, Nebraska, during the first five years (2004-2009) of this initiative. The report summarizes treatment system operational data and regulatory compliance monitoring results for the site during this period, together with the results of the targeted groundwater sampling and analysis for volatile organic compounds (VOCs) conducted in early 2010 (following completion of the fifth year of systems operation), to assess the initial five years of progress of the Utica remediation effort. On the basis of the 2003 groundwater sampling results, a remedial system employing 4 extraction wells (GWEX1-GWEX4), with groundwater treatment by spray irrigation and conventional air stripping, was implemented with the concurrence of the CCC/USDA and the agencies (Table 1.1). The principal components of the system are shown in Figure 1.3 and are briefly described in Section 1.2. Operation of well GWEX4 and the associated air stripper began on October 29, 2004, and routine operation of wells GWEX1-GWEX3 and the spray irrigation treatment units began on November 22, 2004.

54 ENVIRONMENTAL SCIENCES↗

Corrective Action Decision Document/Corrective Action Plan for Corrective Action Unit 98: Frenchman Flat, Nevada National Security Site, Nevada, Revision 1

This CADD/CAP follows the Corrective Action Investigation (CAI) stage, which results in development of a set of contaminant boundary forecasts produced from groundwater flow and contaminant transport modeling of the Frenchman Flat CAU. The Frenchman Flat CAU is located in the southeastern portion of the NNSS and comprises 10 underground nuclear tests. The tests were conducted between 1965 and 1971 and resulted in the release of radionuclides in the subsurface in the vicinity of the test cavities. Two important aspects of the corrective action process are presented within this CADD/CAP. The CADD portion describes the results of the Frenchman Flat CAU data-collection and modeling activities completed during the CAI stage. The corrective action objectives and the actions recommended to meet the objectives are also described. The CAP portion describes the corrective action implementation plan. The CAP begins with the presentation of CAU regulatory boundary objectives and initial use restriction boundaries that are identified and negotiated by NNSA/NSO and the Nevada Division of Environmental Protection (NDEP). The CAP also presents the model evaluation process designed to build confidence that the flow and contaminant transport modeling results can be used for the regulatory decisions required for CAU closure. The first two stages of the strategy have been completed for the Frenchman Flat CAU. A value of information analysis and a CAIP were developed during the CAIP stage. During the CAI stage, a CAIP addendum was developed, and the activities proposed in the CAIP and addendum were completed. These activities included hydrogeologic investigation of the underground testing areas, aquifer testing, isotopic and geochemistry-based investigations, and integrated geophysical investigations. After these investigations, a groundwater flow and contaminant transport model was developed to forecast contaminant boundaries that enclose areas potentially exceeding the Safe Drinking Water Act radiological standards at any time within 1,000 years. An external peer review of the groundwater flow and contaminant transport model was completed, and the model was accepted by NDEP to allow advancement to the CADD/CAP stage. The CADD/CAP stage focuses on model evaluation to ensure that existing models provide adequate guidance for the regulatory decisions regarding monitoring and institutional controls. Data-collection activities are identified and implemented to address key uncertainties in the flow and contaminant transport models. During the CR stage, final use restriction boundaries and CAU regulatory boundaries are negotiated and established; a long-term closure monitoring program is developed and implemented; and the approaches and policies for institutional controls are initiated. The model evaluation process described in this plan consists of an iterative series of five steps designed to build confidence in the site conceptual model and model forecasts. These steps are designed to identify data-collection activities (Step 1), document the data-collection activities in the 0CADD/CAP (Step 2), and perform the activities (Step 3). The new data are then assessed; the model is refined, if necessary; the modeling results are evaluated; and a model evaluation report is prepared (Step 4). The assessments are made by the modeling team and presented to the pre-emptive review committee. The decision is made by the modeling team with the assistance of the pre-emptive review committee and concurrence of NNSA/NSO to continue data and model assessment/refinement, recommend additional data collection, or recommend advancing to the CR stage. A recommendation to advance to the CR stage is based on whether the model is considered to be sufficiently reliable for designing a monitoring system and developing effective institutional controls. The decision to advance to the CR stage or to return to step 1 of the process is then made by NDEP (Step 5).

54 ENVIRONMENTAL SCIENCES↗

Compilation of PRF Canyon Floor Pan Sample Analysis Results

On September 28, 2015, debris collected from the PRF (236-Z) canyon floor, Pan J, was observed to exhibit chemical reaction. The material had been transferred from the floor pan to a collection tray inside the canyon the previous Friday. Work in the canyon was stopped to allow Industrial Hygiene to perform monitoring of the material reaction. Canyon floor debris that had been sealed out was sequestered at the facility, a recovery plan was developed, and drum inspections were initiated to verify no additional reactions had occurred. On October 13, in-process drums containing other Pan J material were inspected and showed some indication of chemical reaction, limited to discoloration and degradation of inner plastic bags. All Pan J material was sealed back into the canyon and returned to collection trays. Based on the high airborne levels in the canyon during physical debris removal, ETGS (Encapsulation Technology Glycerin Solution) was used as a fogging/lock-down agent. On October 15, subject matter experts confirmed a reaction had occurred between nitrates (both Plutonium Nitrate and Aluminum Nitrate Nonahydrate (ANN) are present) in the Pan J material and the ETGS fixative used to lower airborne radioactivity levels during debris removal. Management stopped the use of fogging/lock-down agents containing glycerin on bulk materials, declared a Management Concern, and initiated the Potential Inadequacy in the Safety Analysis determination process. Additional drum inspections and laboratory analysis of both reacted and unreacted material are planned. This report compiles the results of many different sample analyses conducted by the Pacific Northwest National Laboratory on samples collected from the Plutonium Reclamation Facility (PRF) floor pans by the CH2MHill’s Plateau Remediation Company (CHPRC). Revision 1 added Appendix G that reports the results of the Gas Generation Rate and methodology. The scope of analyses requested by CHPRC includes the determination of common anions, gamma spectrometry, metals, corrosivity, organics and alpha spectrometry (note: alpha spectrometry was cancelled during the performance of this work with concurrence from CHPRC). Results may help elucidate the components that led to the unexpected reaction in the canyon as well as inform the radiological and hazardous characteristics. The specific anions, gamma emitters, organics and metals requested by CHPRC are provided in the analytical reports sections. The individual analyses were conducted under the Plutonium Finishing Plant (PFP) Floor Pan Evaluation Project Quality Assurance Project Plan (PFP Floor Pan Evaluation QAPP, Revision 0.) developed by PNNL specifically for this project. The final reports for each analysis set are included in this compilation of the results. Each package was reviewed under the PFP Floor Pan Evaluation Project Quality Assurance Project Plan so no additional reviews were conducted in this compilation task. The Gas Generation Rates in Appendix G were conducted under the PNNL “How Do I…” quality assurance program and were NOT conducted under the Plutonium Finishing Plant (PFP) Floor Pan Evaluation Project Quality Assurance Project Plan (PFP Floor Pan Evaluation QAPP, Revision 0.).

36 MATERIALS SCIENCE↗

Aircraft Cabin Environmental Quality Sensors

The Indoor Environment Department at Lawrence Berkeley National Laboratory (LBNL) teamed with seven universities to participate in a Federal Aviation Administration (FAA) Center of Excellence (COE) for research on environmental quality in aircraft. This report describes research performed at LBNL on selecting and evaluating sensors for monitoring environmental quality in aircraft cabins, as part of Project 7 of the FAA's COE for Airliner Cabin Environmental Research (ACER)1 effort. This part of Project 7 links to the ozone, pesticide, and incident projects for data collection and monitoring and is a component of a broader research effort on sensors by ACER. Results from UCB and LBNL's concurrent research on ozone (ACER Project 1) are found in Weschler et al., 2007; Bhangar et al. 2008; Coleman et al., 2008 and Strom-Tejsen et al., 2008. LBNL's research on pesticides (ACER Project 2) in airliner cabins is described in Maddalena and McKone (2008). This report focused on the sensors needed for normal contaminants and conditions in aircraft. The results are intended to complement and coordinate with results from other ACER members who concentrated primarily on (a) sensors for chemical and biological pollutants that might be released intentionally in aircraft; (b) integration of sensor systems; and (c) optimal location of sensors within aircraft. The parameters and sensors were selected primarily to satisfy routine monitoring needs for contaminants and conditions that commonly occur in aircraft. However, such sensor systems can also be incorporated into research programs on environmental quality in aircraft cabins.

42 ENGINEERING↗

Parallel computing for power system climate resiliency: Solving a large-scale stochastic capacity expansion problem with mpi-sppy

Here we propose a nodal stochastic generation and transmission expansion planning model that incorporates the output from high-resolution global climate models through load and generation availability scenarios. We implement our model in Pyomo and perform computational studies on a realistically-sized test case of the California electric grid in a high performance computing environment. We propose model reformulations and algorithm tuning to efficiently solve this large problem using a variant of the Progressive Hedging Algorithm. We utilize the parallelization capabilities and overall versatility of mpi-sppy, exploiting its hub-and-spoke architecture to concurrently obtain inner and outer bounds on an optimal expansion plan. Initial results show that instances with 360 representative days on a system with over 8,000 buses can be solved to within 5% of optimality in under 4 h of wall clock time, a first step towards solving a large-scale power system expansion planning problem across a wide range of climate-informed operational scenarios.

24 POWER TRANSMISSION AND DISTRIBUTION↗

SU-F-T-167: Measurement of Proton Beam Spot Size Using a 2D Ion Chamber Array for Daily QA

Purpose: To develop an algorithm which can calculate the Full Width Half Maximum (FWHM) of a Proton Pencil Beam from a 2D dimensional ion chamber array (IBA Matrixx) with limited spatial resolution ( 7.6 mm inter chamber distance). The algorithm would allow beam FWHM measurements to be taken during daily QA without an appreciable time increase. Methods: Combinations of 147 MeV single spot beams were delivered onto an IBA Matrixx and concurrently on EBT3 films for a standard. Data were collected around the Bragg Peak region and evaluated by a custom MATLAB script based on our algorithm using a least squared analysis. A set of artificial data, modified with random noise, was also processed to test for robustness. Results: The Matlab script processed Matixx data shows acceptable agreement (within 5%) with film measurements with no single measurement differing by more than 1.8 mm. In cases where the spots show some degree of asymmetry, the algorithm is able to resolve the differences. The algorithm was able to process artificial data with noise up to 15% of the maximum value. Time assays of each measurement took less than 3 minutes to perform, indicating that such measurements may be efficiently added to daily QA treatment. Conclusion: The developed algorithm can be implemented in daily QA program for Proton Pencil Beam scanning beams (PBS) with Matrixx to extract spot size and position information. The developed algorithm may be extended to small field sizes in photon clinic.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Computer-based procedure for field activities: Results from three evaluations at nuclear power plants

Nearly all activities that involve human interaction with the systems of a nuclear power plant are guided by procedures. The paper-based procedures (PBPs) currently used by industry have a demonstrated history of ensuring safety; however, improving procedure use could yield tremendous savings in increased efficiency and safety. One potential way to improve procedure-based activities is through the use of computer-based procedures (CBPs). Computer-based procedures provide the opportunity to incorporate context driven job aids, such as drawings, photos, just-in-time training, etc into CBP system. One obvious advantage of this capability is reducing the time spent tracking down the applicable documentation. Additionally, human performance tools can be integrated in the CBP system in such way that helps the worker focus on the task rather than the tools. Some tools can be completely incorporated into the CBP system, such as pre-job briefs, placekeeping, correct component verification, and peer checks. Other tools can be partly integrated in a fashion that reduces the time and labor required, such as concurrent and independent verification. Another benefit of CBPs compared to PBPs is dynamic procedure presentation. PBPs are static documents which limits the degree to which the information presented can be tailored to the task and conditions when the procedure is executed. The CBP system could be configured to display only the relevant steps based on operating mode, plant status, and the task at hand. A dynamic presentation of the procedure (also known as context-sensitive procedures) will guide the user down the path of relevant steps based on the current conditions. This feature will reduce the user’s workload and inherently reduce the risk of incorrectly marking a step as not applicable and the risk of incorrectly performing a step that should be marked as not applicable. As part of the Department of Energy’s (DOE) Light Water Reactors Sustainability Program, researchers at Idaho National Laboratory (INL) along with partners from the nuclear industry have been investigating the design requirements for computer-based work instructions (including operations procedures, work orders, maintenance procedures, etc.) to increase efficiency, safety, and cost competitiveness of existing light water reactors.

99 GENERAL AND MISCELLANEOUS↗

DIORAMA Communications

Diorama is written as a collection of modules that can run in separate threads or in separate processes. This defines a clear interface between the modules and also allows concurrent processing of different parts of the pipeline. The pipeline is determined by a description in a scenario file[Norman and Tornga, 2012, Tornga and Norman, 2014]. The scenario manager parses the XML scenario and sets up the sequence of modules which will generate an event, propagate the signal to a set of sensors, and then run processing modules on the results provided by those sensor simulations. During a run a variety of “observer” and “processor” modules can be invoked to do interim analysis of results. Observers do not modify the simulation results, while processors may affect the final result. At the end of a run results are collated and final reports are put out. A detailed description of the scenario file and how it puts together a simulation are given in [Tornga and Norman, 2014]. The processing pipeline and how to program it with the Diorama API is described in Tornga et al. [2015] and Tornga and Wakeford [2015]. In this report I describe the communications infrastructure that is used.

97 MATHEMATICS AND COMPUTING↗

Risk-based Prioritization of Facility Decommissioning and Environmental Restoration Projects in the National Nuclear Legacy Liabilities Program at the Chalk River Laboratory - 13564

Chalk River Laboratory (CRL), located in Ontario Canada, has a large number of remediation projects currently in the Nuclear Legacy Liabilities Program (NLLP), including hundreds of facility decommissioning projects and over one hundred environmental remediation projects, all to be executed over the next 70 years. Atomic Energy of Canada Limited (AECL) utilized WorleyParsons to prioritize the NLLP projects at the CRL through a risk-based prioritization and ranking process, using the WorleyParsons Sequencing Unit Prioritization and Estimating Risk Model (SUPERmodel). The prioritization project made use of the SUPERmodel which has been previously used for other large-scale site prioritization and sequencing of facilities at nuclear laboratories in the United States. The process included development and vetting of risk parameter matrices as well as confirmation/validation of project risks. Detailed sensitivity studies were also conducted to understand the impacts that risk parameter weighting and scoring had on prioritization. The repeatable prioritization process yielded an objective, risk-based and technically defendable process for prioritization that gained concurrence from all stakeholders, including Natural Resources Canada (NRCan) who is responsible for the oversight of the NLLP. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

MONOLITHIC FUEL FABRICATION PROCESS DEVELOPMENT AT THE IDAHO NATIONAL LABORATORY

Within the Reduced Enrichment for Research and Test Reactors (RERTR) program directed by the US Department of Energy (DOE), UMo fuel-foils are being developed in an effort to realize high density monolithic fuel plates for use in high-flux research and test reactors. Namely, targeted are reactors that are not amenable to Low Enriched Uranium (LEU) fuel conversion via utilization of high density dispersion-based fuels, i.e. 8-9 gU/cc. LEU conversion of reactors having a need for >8-9 gU/cc fuel density will only be possible by way of monolithic fuel forms. The UMo fuel foils under development afford fuel meat density of ~16 gU/cc and thus have the potential to facilitate LEU conversions without any significant reactor-performance penalty. Two primary challenges have been established with respect to UMo monolithic fuel development; namely, fuel element fabrication and in-reactor fuel element performance. Both issues are being addressed concurrently at the Idaho National Laboratory. An overview is provided of the ongoing monolithic UMo fuel development effort at the Idaho National Laboratory (INL); including development of complex/graded fuel foils. Fabrication processes to be discussed include: UMo alloying and casting, foil fabrication via hot rolling, fuel-clad interlayer application via co-rolling and thermal spray processes, clad bonding via Hot Isostatic Pressing (HIP) and Friction Bonding (FB), and fuel plate finishing.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)

The development of modern processors exhibits two trends that complicate the optimizations of modern software. The first is the increasing sensitivity of processors' throughput to irregularities in computation. With more processors produced through a massive integration of simple cores, future systems will increasingly favor regular data-level parallel computations, but deviate from the needs of applications with complex patterns. Some evidences are already shown on Graphic Processing Units (GPU): Irregular data accesses (e.g., indirect references A[D[i]]) and conditional branches are limiting many GPU applications' performance at a level an order of magnitude lower than the peak of GPU. The second hardware trend is the growing gap between memory bandwidth and the aggregate speed—that is, the sum of all cores' computing power—of a Chip Multiprocessor (CMP). Despite the capped growth of the peak CPU speed, the aggregate speed of a CMP keeps increasing as more cores get into a single chip. It is expected that by 2018, node concurrency in an exascale system will increase by hundreds of times, whereas, memory bandwidth will expand by only 10 to 20 times. Consequently, data movement and storage is expected to consume more than 70% of the total system power. Bridging this gap is difficult; the complexities of modern CMP memory hierarchy make it even harder: Data cache becomes shared among computing units, and the sharing is often non-uniform—whether two computing units share a cache depends on their proximity and the level of the cache. On the recent IBM Power7 architecture, for instance, four hardware contexts (or SMT threads) in a core share the entire memory hierarchy, all cores in one chip share an on-chip L3 cache, and cores across chips share L3 and main memory through off-chip connections. These two trends complicate the translation of computing power into performance, especially for a program with either intensive data accesses or complex patterns in data accesses or control flow paths. Unfortunately, both attributes present and will persist in a class of important applications. For instance, many scientific simulations deal with a large volume of data. And meanwhile, as most real-world processes are non-uniform and evolving (e.g., the evolution of a galaxy or the process of a drug injection), both the computations and data accesses of these programs tend to be irregular and dynamically changing. Currently, the lack of support to these applications on modern CMP severely limits their performance. On GPU, as our recent study shows and other studies echo, performance enhancement of a factor of integers is possible when memory accesses or control flows are streamlined for a set of GPU applications. On multicore CPU, our studies show that traditional locality enhancement, for being oblivious to the new features of multicore memory hierarchy, may even cause large slowdown to data-intensive dynamic applications. The severity of the issues is expected to worsen as the two hardware trends continue. Some recent studies try to match software with the trends, but in a limited scope or manner. For irregularities on GPU, most studies focus on irregularities analyzable through static analysis (e.g., data accesses in regular loops). Dynamic irregularities are harder to address because the needed analysis and transformations typically have to happen at run time. Some other research resorts to hardware extensions, an actual adoption of which is unclear for the entailed space cost and complexity. For data locality, recent years have seen some exploitations of the new memory hierarchy on multicore for performance, but most of them are on process or thread scheduling, rather than program transformations. Our recent study reveals that program-level transformations may magnify the scheduling benefits by a factor of seven, concluding that program-level transformation should play a central role for data locality enhancement on modern CMP. But research in this direction has been sparse, and most have focused on data layout or cache performance modeling, rather than program transformations to match with the new memory hierarchy features. Overall, it is still an open question how to bridge the gap between dynamic computations and the two prominent properties of modern processors. The goal of this project is to develop a set of techniques and software tools to enhance the matching between memory accesses in dynamic simulations and the prominent features of modern and future CMP, alleviating the memory performance issues for petascale and exascale computing. This report summarizes the discoveries and products produced throughout this project. It includes free launch, a new software approach to overcoming the shortcomings of both methods; coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution; algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Supporting dynamic parallelism is important for GPU to benefit a broad range of applications. There are currently two fundamental ways for programs to exploit dynamic parallelism on GPU: a software-based approach with software-managed worklists, and a hardware-based approach through dynamic subkernel launches. Neither is satisfactory. The former is complicated to program and is often subject to some load imbalance; the latter suffers large runtime overhead. In this work, we propose free launch, a new software approach to overcoming the shortcomings of both methods. It allows programmers to use subkernel launches to express dynamic parallelism. It employs a novel compiler-based code transformation named subkernel launch removal to replace the subkernel launches with the reuse of parent threads. Coupled with an adaptive task assignment mechanism, the transformation reassigns the tasks in the subkernels to the parent threads with a good load balance. The technique requires no hardware extensions, immediately deployable on existing GPUs. It keeps the programming convenience of the subkernel launch-based approach while avoiding its large runtime overhead. Meanwhile, its superior load balancing makes it outperform manual worklist-based techniques by 3X on average. The work was published at the 48th Annual IEEE/ACM International Symposium on Microarchitecture (Micro'2015). A Graphic Processing Unit (GPU) system is typically equipped with many types of memory (e.g., global, constant, texture, shared, cache). Data placement determines what data are placed on which type of memory, essential for GPU memory performance. Prior optimizations of data placement always require a single view of a data object on memory, which limits the optimization effectiveness. In this work, we propose coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution. We demonstrate that under certain conditions, the multiple views can remain incoherent while facilitating enhanced data placement. We present a theorem and some compiler support to ensure the soundness of the usage of coherence-free multiview. We further develop reference-discerning data placement, a new way to enhance data placements on GPU. It enables more flexible data placements by using coherence-free multiview to leverage the slack in coherence requirement of some GPU programs. Experiments on three types of GPU systems show that, with less than 200KB space cost, the new data placement technique can provide a 1.6X average (up to 4.27X) speedup. In addition, we have examined algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Computing distances among data points is an essential part of many important algorithms in data analytics, graph analysis, and other domains. In each of these domains, developers have spent significant manual effort optimizing algorithms, often through novel applications of the triangle equality, in order to minimize the number of distance computations in the algorithms. In this work, we observe that many algorithms across these domains can be generalized as an instance of a generic distance-related abstraction. Based on this abstraction, we derive seven principles for correctly applying the triangular inequality to optimize distance-related algorithms. Guided by the findings, we develop Triangular OPtimizer (TOP), the first software framework that is able to automatically produce optimized algorithms that either matches or outperforms manually designed algorithms for solving distance-related problems. TOP achieves up to 237x speedups and 2.5X on average. The work has been published at the 32nd International Conference on Machine Learning (ICML'15) and the 41st International Conference on Very Large Data Bases (VLDB'15).

42 ENGINEERING↗

Glprof: A Gprof inspired, Callgraph-oriented Per-Object Disseminating Memory Access Multi-Cache Profiler

Application analysis is facilitated through a number of program profiling tools. The tools vary in their complexity, ease of deployment, design, and profiling detail. Specifically, understand- ing, analyzing, and optimizing is of particular importance for scientific applications where minor changes in code paths and data-structure layout can have profound effects. Understanding how intricate data-structures are accessed and how a given memory system responds is a complex task. In this paper we describe a trace profiling tool, Glprof, specifically aimed to lessen the burden of the programmer to pin-point heavily involved data-structures during an application's run-time, and understand data-structure run-time usage. Moreover, we showcase the tool's modularity using additional cache simulation components. We elaborate on the tool's design, and features. Finally we demonstrate the application of our tool in the context of Spec bench- marks using the Glprof profiler and two concurrently running cache simulators, PPC440 and AMD Interlagos.

Janjusic, Tommy↗

Regulatory Considerations in the Development of Radiation-Drug Combinations

Radiation therapy remains a fundamental treatment for patients with cancer. Despite an increasing number of targeted molecular therapies that are US Food and Drug Administration (FDA)-approved for the treatment of patients with metastatic disease, there has been very little progress made in terms of drugs used concurrently with radiation. This article reviews the existing regulatory framework in which cancer drugs may be developed for use in combination with radiation therapy from the perspective of the FDA. To briefly summarize: (1) nonclinical studies are a critical first step to ensure that drugs are safe for use in humans; however, additional nonclinical studies of a drug with radiation may not be required before a clinical trial in combination with radiation as long as the safety profile of the drug has been characterized in humans. The FDA determines the quality of evidence required before studying a drug in combination with radiation on a case-by-case basis. (2) Although often impractical to consider late toxicities during dose-escalation, late adverse events should be captured and taken into consideration when determining the final dose and schedule to take forward during drug development. (3) There are a number of expedited programs for cancer drug development, including accelerated approval, a conditional approval that allows for use of earlier clinical endpoints when the data suggests a clinically meaningful improvement over available therapy. (4) The Agency encourages sponsors to discuss their development plan with the appropriate FDA review division in formal regulatory meetings.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

CUDA Computation of the Feynman Distribution

In 2006, NVIDIA Corporation introduced the Computer Unified Device Architecture (CUDA) for C, C++, and Fortran programming languages. The CUDA parallel computing platform uses the Graphics Processing Unit (GPU) of NVIDIA graphics card, and it is designed for parallel computations. In these computations, the same algorithms are executed on the data in parallel. The GPU contains many Streaming Multiprocessors (SM) that can independently execute blocks of threads. A thread is a sequence of serial programmed instructions. The SM executes threads in groups of 32 and a group of 32 threads is called warp. The threads within a block can be synchronized and have access to a shared memory that is visible only to them. Conversely, blocks of threads cannot be synchronized and have access only to the global memory of the graphics card. This memory is called device memory and is independent of the host memory controlled by the Central Processing Unit (CPU). A single thread has access to a local memory (register) that is not visible to any other thread. The part of the program that is executed on the graphics card is called kernel. Each thread of the kernel has a unique identification index that is accessible by the in-built variable threadIdx; the latter is a 3-component vector. A threads block can contain up to 1024 threads. Blocks of threads have a unique identification index that is accessible by the in-built variable blockIdx; the latter is a 3-component vector. The in-built variable blockDim contains the number of threads per block; this variable is a 3-component vector. Threads and blocks represent different levels of parallelism; a kernel is executed as a grid of blocks of threads. Figure 1 illustrates a grid of thread blocks for a GPU with 2 SM. A SM can execute multiple blocks of threads in parallel; a block of threads is executed on a single SM. A GPU can execute multiple kernels concurrently. This work has used the Quadro M5000 graphics card with the following technical specifications: - 2048 cores with 861 MHz frequency, - 16 streaming multiprocessors, - 5.2 trillion transistors, - 8.2 Gb device memory, - 64 kb register memory (per thread), - 48 kb shared memory (per block) - 211.6 GB/s memory bandwidth, and - 3.5 GFLOPS floating point performance.

97 MATHEMATICS AND COMPUTING↗

CO{sub 2} FORMATION IN QUIESCENT CLOUDS: AN EXPERIMENTAL STUDY OF THE CO + OH PATHWAY

The formation of CO{sub 2} in quiescent regions of molecular clouds is not yet fully understood, despite CO{sub 2} having an abundance of around 10%-34% H{sub 2}O. We present a study of the formation of CO{sub 2} via the nonenergetic route CO + OH on nonporous H{sub 2}O and amorphous silicate surfaces. Our results are in the form of temperature-programmed desorption spectra of CO{sub 2} produced via two experimental routes: O{sub 2} + CO + H and O{sub 3} + CO + H. The maximum yield of CO{sub 2} is around 8% with respect to the starting quantity of CO, suggesting a barrier to CO + OH. The rate of reaction, based on modeling results, is 24 times slower than O{sub 2} + H. Our model suggests that competition between CO{sub 2} formation via CO + OH and other surface reactions of OH is a key factor in the low yields of CO{sub 2} obtained experimentally, with relative reaction rates of k{sub CO+H}<<k{sub CO+OH}< k{sub H{sub 2}O{sub 2}+H}< k{sub OH+H},k{sub O{sub 2}+H}. Astrophysically, the presence of CO{sub 2} in low A{sub V} regions of molecular clouds could be explained by the reaction CO + OH occurring concurrently with the formation of H{sub 2}O via the route OH + H.

79 ASTRONOMY AND ASTROPHYSICS↗

NETL to establish Dynamic Simulation Research and Training Center to promote IGCC technology with CO2 cpture

To meet increasing demand for education and experience with commercial-scale, coal-fired, integrated gasification combined cycle (IGCC) plants with CO2 capture, the Department of Energy’s (DOE) National Energy Technology Laboratory (NETL) is leading a project to deploy a generic, full-scope, real-time IGCC dynamic plant simulator for use in establishing a world-class research and training center, and to promote and demonstrate IGCC technology to power industry personnel. The simulator, being built by Invensys Process Systems (IPS), will be installed at two separate sites, at NETL and West Virginia University (WVU), and will combine a process/gasification simulator with a power/combined-cycle simulator together in a single dynamic simulation framework for use in engineering research studies and training applications. The simulator, scheduled to be launched in mid-year 2010, will have the following capabilities: High-fidelity, dynamic model of process-side (gasification and gas cleaning with CO2 capture) and power-block-side (combined cycle) for a generic IGCC plant fueled by coal and/or petroleum coke. Highly flexible configuration that allows concurrent training on separate gasification and combined cycle simulators, or up to two IGCC simulators. Ability to enhance and modify the plant model to facilitate studies of changes in plant configuration, equipment, and control strategies to support future R&D efforts. Training capabilities including startup, shutdown, load following and shedding, response to fuel and ambient condition variations, control strategy analysis (turbine vs. gasifier lead, etc.), representative malfunctions/trips, alarms, scenarios, trending, snapshots, data historian, etc. To support this effort, process descriptions and control strategies were developed for key sections of the plant as part of the detailed functional specification, which is serving as the basis of the simulator development. In this paper, we highlight the contents of the detailed functional specification for the simulator. We also describe the engineering, design, and expert testing process that the simulator will undergo in order to ensure that maximum fidelity is built into the generic simulator. Future applications and training programs associated with gasification, combined cycle, and IGCC simulations are discussed, including plant operation and control demonstrations, as well as education and training services.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗