Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “high performance computing systems”

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 1,747 records · Page 97

Performance of a high-sensitivity dedicated cardiac SPECT scanner for striatal uptake quantification in the brain based on analysis of projection data

Purpose: The authors have previously reported the advantages of high-sensitivity single-photon emission computed tomography (SPECT) systems for imaging structures located deep inside the brain. DaTscan (Isoflupane I-123) is a dopamine transporter (DaT) imaging agent that has shown potential for early detection of Parkinson disease (PD), as well as for monitoring progression of the disease. Realizing the full potential of DaTscan requires efficient estimation of striatal uptake from SPECT images. They have evaluated two SPECT systems, a conventional dual-head gamma camera with low-energy high-resolution collimators (conventional) and a dedicated high-sensitivity multidetector cardiac imaging system (dedicated) for imaging tasks related to PD. Methods: Cramer-Rao bounds (CRB) on precision of estimates of striatal and background activity concentrations were calculated from high-count, separate acquisitions of the compartments (right striata, left striata, background) of a striatal phantom. CRB on striatal and background activity concentration were calculated from essentially noise-free projection datasets, synthesized by scaling and summing the compartment projection datasets, for a range of total detected counts. They also calculated variances of estimates of specific-to-nonspecific binding ratios (BR) and asymmetry indices from these values using propagation of error analysis, as well as the precision of measuring changes in BR on the order of the average annual decline in early PD. Results: Under typical clinical conditions, the conventional camera detected 2 M counts while the dedicated camera detected 12 M counts. Assuming a normal BR of 5, the standard deviation of BR estimates was 0.042 and 0.021 for the conventional and dedicated system, respectively. For an 8% decrease to BR = 4.6, the signal-to-noise ratio were 6.8 (conventional) and 13.3 (dedicated); for a 5% decrease, they were 4.2 (conventional) and 8.3 (dedicated). Conclusions: This implies that PD can be detected earlier with the dedicated system than with the conventional system; therefore, earlier identification of PD progression should be possible with the high-sensitivity dedicated SPECT camera.

60 APPLIED LIFE SCIENCES↗

Evaluation of CHO Benchmarks on the Arria 10 FPGA using Intel FPGA SDK for OpenCL

The OpenCL standard is an open programming model for accelerating algorithms on heterogeneous computing system. OpenCL extends the C-based programming language for developing portable codes on different platforms such as CPU, Graphics processing units (GPUs), Digital Signal Processors (DSPs) and Field Programmable Gate Arrays (FPGAs). The Intel FPGA SDK for OpenCL is a suite of tools that allows developers to abstract away the complex FPGA-based development flow for a high-level software development flow. Users can focus on the design of hardware-accelerated kernel functions in OpenCL and then direct the tools to generate the low-level FPGA implementations. The approach makes the FPGA-based development more accessible to software users as the needs for hybrid computing using CPUs and FPGAs are increasing. It can also significantly reduce the hardware development time as users can evaluate different ideas with high-level language without deep FPGA domain knowledge. Benchmarking of OpenCL-based framework is an effective way for analyzing the performance of system by studying the execution of the benchmark applications. CHO is a suite of benchmark applications that provides support for OpenCL [1]. The authors presented CHO as an OpenCL port of the CHStone benchmark. Using Altera OpenCL (AOCL) compiler to synthesize the benchmark applications, they listed the resource usage and performance of each kernel that can be successfully synthesized by the compiler. In this report, we evaluate the resource usage and performance of the CHO benchmark applications using the Intel FPGA SDK for OpenCL and Nallatech 385A FPGA board that features an Arria 10 FPGA device. The focus of the report is to have a better understanding of the resource usage and performance of the kernel implementations using Arria-10 FPGA devices compared to Stratix-5 FPGA devices. In addition, we also gain knowledge about the limitations of the current compiler when it fails to synthesize a benchmark application.

97 MATHEMATICS AND COMPUTING↗

High Performance, Three-Dimensional Bilateral Filtering

Image smoothing is a fundamental operation in computer vision and image processing. This work has two main thrusts: (1) implementation of a bilateral filter suitable for use in smoothing, or denoising, 3D volumetric data; (2) implementation of the 3D bilateral filter in three different parallelization models, along with parallel performance studies on two modern HPC architectures. Our bilateral filter formulation is based upon the work of Tomasi [11], but extended to 3D for use on volumetric data. Our three parallel implementations use POSIX threads, the Message Passing Interface (MPI), and Unified Parallel C (UPC), a Partitioned Global Address Space (PGAS) language. Our parallel performance studies, which were conducted on a Cray XT4 supercomputer and aquad-socket, quad-core Opteron workstation, show our algorithm to have near-perfect scalability up to 120 processors. Parallel algorithms, such as the one we present here, will have an increasingly important role for use in production visual analysis systems as the underlying computational platforms transition from single- to multi-core architectures in the future.

97 MATHEMATICS AND COMPUTING↗

Effects of increasing tip velocity on wind turbine rotor design.

A reduction in cost of energy from wind is anticipated when maximum allowable tip velocity is allowed to increase. Rotor torque decreases as tip velocity increases and rotor size and power rating are held constant. Reduction in rotor torque yields a lighter weight gearbox, a decrease in the turbine cost, and an increase in the capacity for the turbine to deliver cost competitive electricity. The high speed rotor incurs costs attributable to rotor aero-acoustics and system loads. The increased loads of high speed rotors drive the sizing and cost of other components in the system. Rotor, drivetrain, and tower designs at 80 m/s maximum tip velocity and 100 m/s maximum tip velocity are created to quantify these effects. Component costs, annualized energy production, and cost of energy are computed for each design to quantify the change in overall cost of energy resulting from the increase in turbine tip velocity. High fidelity physics based models rather than cost and scaling models are used to perform the work. Results provide a quantitative assessment of anticipated costs and benefits for high speed rotors. Finally, important lessons regarding full system optimization of wind turbines are documented.

17 WIND ENERGY↗

Fast equilibration protocol for million atom systems of highly entangled linear polyethylene chains

Equilibrated systems of entangled polymer melts cannot be produced using direct brute force equilibration due to the slow reptation dynamics exhibited by high molecular weight chains. Instead, these dense systems are produced using computational techniques such as Monte Carlo-Molecular Dynamics hybrid algorithms, though the use of soft potentials has also shown promise mainly for coarse-grained polymeric systems. Through the use of soft-potentials, the melt can be equilibrated via molecular dynamics at intermediate and long length scales prior to switching to a Lennard-Jones potential. We will outline two different equilibration protocols, which use various degrees of information to produce the starting configurations. In one protocol, we use only the equilibrium bond angle, bond length, and target density during the construction of the simulation cell, where the information is obtained from available experimental data and extracted from the force field without performing any prior simulation. In the second protocol, we moreover utilize the equilibrium radial distribution function and dihedral angle distribution. This information can be obtained from experimental data or from a simulation of short unentangled chains. Both methods can be used to prepare equilibrated and highly entangled systems, but the second protocol is much more computationally efficient. These systems can be strictly monodisperse or optionally polydisperse depending on the starting chain distribution. Our protocols, which utilize a soft-core harmonic potential, will be applied for the first time to equilibrate a million particle system of polyethylene chains consisting of 1000 united atoms at various temperatures. Calculations of structural and entanglement properties demonstrate that this method can be used as an alternative towards the generation of entangled equilibrium structures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Computational discovery of stable and metastable ternary oxynitrides

Materials design from first principles enables exploration of uncharted chemical spaces. Extensive computational searches have been performed for mixed-cation ternary compounds, but mixed-anion systems are gaining increased interest as well. Central to computational discovery is the crystal structure prediction, where the trade-off between reliance on prototype structures and size limitations of unconstrained sampling has to be navigated. We approach this challenge by letting two complementary structure sampling approaches compete. We use the kinetically limited minimization approach for high-throughput unconstrained crystal structure prediction in smaller cells up to 21 atoms. On the other hand, ternary—and, more generally, multinary—systems often assume structures formed by atomic ordering on a lattice derived from a binary parent structure. Thus, we additionally sample atomic configurations on prototype lattices with cells up to 56 atoms. Using this approach, we searched 65 different charge-balanced oxide–nitride stoichiometries, including six known systems as the control sample. The convex hull analysis is performed both for the thermodynamic limit and for the case of synthesis with activated nitrogen sources. We identified 34 phases that are either on the convex hull or within a viable energy window for potentially metastable phases. We further performed structure sampling for “missing” binary nitrides whose energies are needed for the convex hull analysis. Among these, we discovered metastable Ce3N4 as a nitride analog of the tetravalent cerium oxide, which becomes stable under slightly activated nitrogen condition µN > +0.07 eV. Given the outsize role of CeO2 in research and application, Ce3N4 is a potentially important discovery.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Large Scale Computing and Storage Requirements for Basic Energy Sciences Research

The National Energy Research Scientific Computing Center (NERSC) is the leading scientific computing facility supporting research within the Department of Energy's Office of Science. NERSC provides high-performance computing (HPC) resources to approximately 4,000 researchers working on about 400 projects. In addition to hosting large-scale computing facilities, NERSC provides the support and expertise scientists need to effectively and efficiently use HPC systems. In February 2010, NERSC, DOE's Office of Advanced Scientific Computing Research (ASCR) and DOE's Office of Basic Energy Sciences (BES) held a workshop to characterize HPC requirements for BES research through 2013. The workshop was part of NERSC's legacy of anticipating users future needs and deploying the necessary resources to meet these demands. Workshop participants reached a consensus on several key findings, in addition to achieving the workshop's goal of collecting and characterizing computing requirements. The key requirements for scientists conducting research in BES are: (1) Larger allocations of computational resources; (2) Continued support for standard application software packages; (3) Adequate job turnaround time and throughput; and (4) Guidance and support for using future computer architectures. This report expands upon these key points and presents others. Several 'case studies' are included as significant representative samples of the needs of science teams within BES. Research teams scientific goals, computational methods of solution, current and 2013 computing requirements, and special software and support needs are summarized in these case studies. Also included are researchers strategies for computing in the highly parallel, 'multi-core' environment that is expected to dominate HPC architectures over the next few years. NERSC has strategic plans and initiatives already underway that address key workshop findings. This report includes a brief summary of those relevant to issues raised by researchers at the workshop.

36 MATERIALS SCIENCE↗

RELAP5-3D Pre-test Prediction for High Temperature Test Facility Test PG-26

The RELAP5-3D computer code has been used to perform a pre-test prediction for Test PG-26 in the High Temperature Test Facility at Oregon State University. Test PG-26 will simulate a pressurized conduction cooldown transient following a trip of the primary coolant system gas circulator. The facility is initialized with a steady-state power of 700 kW and prototypic core inlet and outlet helium coolant temperatures of 259 and 687°C, respectively. Sensitivity calculations addressed the flow coastdown time, the reactor cavity cooling system flow rate and temperature, and decay power.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Implications of Grain-Scale Mineralogical Heterogeneity for Radionuclide Transport in Fractured Media

The geological disposal of nuclear waste is based on the multi-barrier concept, comprising various engineered and natural barriers, to confine the radioactive waste and isolate it from the biosphere. Some of the planned repositories for high-level nuclear waste will be hosted in fractured crystalline rock formations. The potential of these formations to act as natural transport barriers is related to two coupled processes: diffusion into the rock matrix and sorption onto the mineral surfaces available in the rock matrix. Different in situ and laboratory experiments have pointed out the ubiquitous heterogeneous nature of the rock matrix: mineral surfaces and pore space are distributed in complex microstructures and their distribution is far from being homogeneous (as typically assumed by Darcy-scale coarse reactive transport models). We use a synthetically generated fracture–matrix system to assess the implications of grain-scale physical and mineralogical heterogeneity on cesium transport and retention. The resulting grain-scale reactive transport model is solved using high-performance computing technologies, and the results are compared with those derived from two alternative models, denoted as upscaled models, where mineral abundance is averaged over the matrix volume. In the grain-scale model, the penetration of cesium into the matrix is faster and the penetration front is uneven and finger-shaped. The analysis of the cesium breakthrough curves computed at two different points in the fracture shows that the upscaled models provide later first-arrival time estimates compared to the grain-scale model. The breakthrough curves computed with the three models converge at late times. These results suggest that spatially averaged upscaled parameters of sorption site distribution can be used to predict the late-time behavior of breakthrough curves but could be inadequate to simulate the early behavior.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Device-scale computational fluid dynamics modeling of carbon dioxide absorption using encapsulated sorbents

Micro-encapsulated carbon dioxide sorbent (MECS) is considered to be a promising material for carbon capture system because of high capture rate, low fabrication cost, and efficient energy consumption. To accelerate the development of MECS technology for industrial deployment, device-scale computational fluid dynamics (CFD) simulations are performed to study hydrodynamics and CO2 absorption behavior in a conceptual fluidized bed reactor. The two-fluid model solver, built in the Multiphase Flow with Interphase eXchanges (MFIX), is adopted to solve the multiphase flow hydrodynamics by representing both fluid and solids as continuous phases. Filtered models are implemented to capture the sub-grid effects that normally cannot be directly considered because MECS particles are too small to be explicitly resolved at such a large scale. Using established CFD tools, this study aims to characterize MECS behavior and predict its carbon capture performance in a conceptual device-scale absorber. The MECS particle size and gas flow rate have significant impact on CO2 capture efficiency. The CO2 capture efficiency decreases with increasing gas flow rate. Smaller particle size lead to better gas-solid homogeneity, as a results CO2 absorption efficiency increases.

Wang, Chao (ORCID:0000000279406614)↗

UPC++: A PGAS extension for C++

Partitioned Global Address Space (PGAS) languages are convenient for expressing algorithms with large, random-access data, and they have proven to provide high performance and scalability through lightweight one-sided communication and locality control. While very convenient for moving data around the system, PGAS languages have taken different views on the model of computation, with the static Single Program Multiple Data (SPMD) model providing the best scalability. In this paper we present UPC++, a PGAS extension for C++ that has three main objectives: 1) to provide an object-oriented PGAS programming model in the context of the popular C++ language, 2) to add useful parallel programming idioms unavailable in UPC, such as asynchronous remote function invocation and multidimensional arrays, to support complex scientific applications, 3) to offer an easy on-ramp to PGAS programming through interoperability with other existing parallel programming systems (e.g., MPI, OpenMP, CUDA). We implement UPC++ with a 'compiler-free' approach using C++ templates and runtime libraries. We borrow heavily from previous PGAS languages and describe the design decisions that led to this particular set of language features, providing significantly more expressiveness than UPC with very similar performance characteristics. We evaluate the programmability and performance of UPC++ using five benchmarks on two representative supercomputers, demonstrating that UPC++ can deliver excellent performance at large scale up to 32K cores while offering PGAS productivity features to C++ applications.

Zheng, Y↗

On the Impact of Execution Models: A Case Study in Computational Chemistry

Efficient utilization of high-performance computing (HPC) platforms is an important and complex problem. Execution models, abstract descriptions of the dynamic runtime behavior of the execution stack, have significant impact on the utilization of HPC systems. Using a computational chemistry kernel as a case study and a wide variety of execution models combined with load balancing techniques, we explore the impact of execution models on the utilization of an HPC system. We demonstrate a 50 percent improvement in performance by using work stealing relative to a more traditional static scheduling approach. We also use a novel semi-matching technique for load balancing that has comparable performance to a traditional hypergraph-based partitioning implementation, which is computationally expensive. Using this study, we found that execution model design choices and assumptions can limit critical optimizations such as global, dynamic load balancing and finding the correct balance between available work units and different system and runtime overheads. With the emergence of multi- and many-core architectures and the consequent growth in the complexity of HPC platforms, we believe that these lessons will be beneficial to researchers tuning diverse applications on modern HPC platforms, especially on emerging dynamic platforms with energy-induced performance variability.

Execution models, computational chemistry, MPI and↗

A Survey Of Techniques for Managing and Leveraging Caches in GPUs

Initially introduced as special-purpose accelerators for graphics applications, graphics processing units (GPUs) have now emerged as general purpose computing platforms for a wide range of applications. To address the requirements of these applications, modern GPUs include sizable hardware-managed caches. However, several factors, such as unique architecture of GPU, rise of CPU–GPU heterogeneous computing, etc., demand effective management of caches to achieve high performance and energy efficiency. Recently, several techniques have been proposed for this purpose. In this paper, we survey several architectural and system-level techniques proposed for managing and leveraging GPU caches. We also discuss the importance and challenges of cache management in GPUs. The aim of this paper is to provide the readers insights into cache management techniques for GPUs and motivate them to propose even better techniques for leveraging the full potential of caches in the GPUs of tomorrow.

cache memory↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, BrianVan↗

Outcomes from the DOE Workshop on Turbulent Flow Simulation at the Exascale

This paper summarizes the outcomes from the Turbulent Flow Simulation at the Exascale: Opportunities and Challenges Workshop, which was held 4-5 August 2015, and was sponsored by the U.S. Department of Energy Office of Advanced Scientific Computing Research. The workshop objective was to define and describe the challenges and opportunities that computing at the exascale will bring to turbulent-flow simulations in applied science and technology. The need for accurate simulation of turbulent flows is evident across the U.S. Department of Energy applied-science and engineering portfolios, including combustion, plasma physics, nuclear-reactor physics, wind energy, and atmospheric science. The workshop brought together experts in turbulent-flow simulation, computational mathematics, and high-performance computing. Building upon previous ASCR workshops on exascale computing, participants defined a research agenda and path forward that will enable scientists and engineers to continually leverage, engage, and direct advances in computational systems on the path to exascale computing.

97 MATHEMATICS AND COMPUTING↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, Brian Van↗

Xenon Recovery at Room Temperature using Metal-Organic Frameworks

Xenon is known to be a very efficient anesthetic gas but its cost prohibits the wider use in medical industry and other potential applications. It has been shown that Xe recovery and recycle from anesthetic gas mixture can significantly reduce its cost as anesthetic. The current technology uses series of adsorbent columns followed by low temperature distillation to recover Xe, which is expensive to use in medical facilities. Herein, we propose much efficient and simpler system to recover and recycle Xe from simulant exhale anesthetic gas mixture at room temperature using metal organic frameworks. Among the MOFs tested, PCN-12 exhibits unprecedented performance with high Xe capacity, Xe/N2 and Xe/O2 selectivity at room temperature. The in-situ synchrotron measurements suggest the Xe is occupied in the small pockets of PCN-12 compared to unsaturated metal centers (UMCs). Computational modeling of adsorption further supports our experimental observation of Xe binding sites in PCN-12.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Communication Requirements and Interconnect Optimization forHigh-End Scientific Applications

The path towards realizing peta-scale computing isincreasingly dependent on building supercomputers with unprecedentednumbers of processors. To prevent the interconnect from dominating theoverall cost of these ultra-scale systems, there is a critical need forhigh-performance network solutions whose costs scale linearly with systemsize. This work makes several unique contributions towards attaining thatgoal. First, we conduct one of the broadest studies to date of high-endapplication communication requirements, whose computational methodsinclude: finite-difference, lattice-bolzmann, particle in cell, sparselinear algebra, particle mesh ewald, and FFT-based solvers. Toefficiently collect this data, we use the IPM (Integrated PerformanceMonitoring) profiling layer to gather detailed messaging statistics withminimal impact to code performance. Using the derived communicationcharacterizations, we next present fit-trees interconnects, a novelapproach for designing network infrastructure at a fraction of thecomponent cost of traditional fat-tree solutions. Finally, we propose theHybrid Flexibly Assignable Switch Topology (HFAST) infrastructure, whichuses both passive (circuit) and active (packet) commodity switchcomponents to dynamically reconfigure interconnects to suit thetopological requirements of scientific applications. Overall ourexploration leads to a promising directions for practically addressingthe interconnect requirements of future peta-scale systems.

99 GENERAL AND MISCELLANEOUS↗