Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallelization efficiency”

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 1,261 records · Page 70

A Framework for Batched and GPU-Resident Factorization Algorithms Applied to Block Householder Transformations

As modern hardware keeps evolving, an increasingly effective approach to developing energy efficient and high-performance solvers is to design them to work on many small size and independent problems. Many applications already need this functionality, especially for GPUs, which are currently known to be about four to five times more energy efficient than multicore CPUs. We describe the development of one-sided factorizations that work for a set of small dense matrices in parallel, and we illustrate our techniques on the QR factorization based on Householder transformations. We refer to this mode of operation as a batched factorization. Our approach is based on representing the algorithms as a sequence of batched BLAS routines for GPU-only execution. This is in contrast to the hybrid CPU-GPU algorithms that rely heavily on using the multicore CPU for specific parts of the workload. But for a system to benefit fully from the GPU's significantly higher energy efficiency, avoiding the use of the multicore CPU must be a primary design goal, so the system can rely more heavily on the more efficient GPU. Additionally, this will result in the removal of the costly CPU-to-GPU communication. Furthermore, we do not use a single symmetric multiprocessor(on the GPU) to factorize a single problem at a time. We illustrate how our performance analysis, and the use of profiling and tracing tools, guided the development and optimization of our batched factorization to achieve up to a 2-fold speedup and a 3-fold energy efficiency improvement compared to our highly optimized batched CPU implementations based on the MKL library(when using two sockets of Intel Sandy Bridge CPUs). Compared to a batched QR factorization featured in the CUBLAS library for GPUs, we achieved up to 5x speedup on the K40 GPU.

Dong, Tingzing Tim↗

Analyzing and Visualizing Cosmological Simulations with ParaView

The advent of large cosmological sky surveys - ushering in the era of precision cosmology - has been accompanied by ever larger cosmological simulations. The analysis of these simulations, which currently encompass tens of billions of particles and up to a trillion particles in the near future, is often as daunting as carrying out the simulations in the first place. Therefore, the development of very efficient analysis tools combining qualitative and quantitative capabilities is a matter of some urgency. In this paper, we introduce new analysis features implemented within ParaView, a fully parallel, open-source visualization toolkit, to analyze large N-body simulations. A major aspect of ParaView is that it can live and operate on the same machines and utilize the same parallel power as the simulation codes themselves. In addition, data movement is in a serious bottleneck now and will become even more of an issue in the future; an interactive visualization and analysis tool that can handle data in situ is fast becoming essential. The new features in ParaView include particle readers and a very efficient halo finder that identifies friends-of-friends halos and determines common halo properties, including spherical overdensity properties. In combination with many other functionalities already existing within ParaView, such as histogram routines or interfaces to programming languages like Python, this enhanced version enables fast, interactive, and convenient analyses of large cosmological simulations. In addition, development paths are available for future extensions.

79 ASTRONOMY AND ASTROPHYSICS↗

Cloud Services Enable Efficient AI-Guided Simulation Workflows across Heterogeneous Resources

Applications which fuse machine learning and simulation are rarely best served by a single computing resource. Highly parallel simulation codes are best deployed on super- computers, while AI tasks used to decide which simulations to perform may be best suited to specialized accelerators. Here we present a Function-as-a-Service (FaaS) system for executing complex, distributed computational campaigns that achieves performance parity with conventional workflow systems without the complexities of secure network connections between compute providers. One innovation enabling high performance is a subsystem that directly moves task data between sites, separate from the cloud-hosted FaaS system used to distribute task instructions. We also introduce a flexible scheduling system that allows us access factor of 2 trade offs between the amount of resources required to solve a problem at each compute site. We anticipate that this system will upgrade multi-site applications from demonstration projects to routine practice in computational science.

Ward, Logan↗

User-Oriented Modeling Tools for Advanced Hybrid and Climate-Appropriate Rooftop Air Conditioners

Hybrid unitary air conditioning systems offer a pathway to substantially reduce energy use and peak electrical demand for cooling, heating, and ventilation in commercial buildings. Hybrid air conditioners incorporate multiple subsystems that are carefully orchestrated to provide climate- and application-specific efficiency advantages. There are a multitude of hybrid system architectures, but common subsystems include: heat recovery ventilation, indirect evaporative cooling, desiccant dehumidification, variable speed fans, modulating dampers, and multi-stage or variable-speed vapor compression cooling. Categorically, hybrid systems can operate in numerous discrete modes. For example: indirect evaporative cooling may operate for periods when the subsystem provides adequate sensible cooling, then vapor compression cooling will be included when more cooling or dehumidification is necessary. Laboratory assessments, field studies, and simulations have demonstrated that hybrid unitary air conditioners could reduce energy use for cooling and ventilation by 30-90% depending on climate and application. Heretofore, it has been challenging - if not impossible - for practitioners to model hybrid air conditioners as part of building energy simulations; and the limitation has severely obstructed broader adoption of technologies in this class. In this project, we developed a new feature for EnergyPlus that enables modeling hybrid unitary air conditioning equipment for building energy simulations. This is a significant advancement for both theory and practice, and confers public benefit by enabling practitioners to evaluate this compelling efficiency technology as a part of building energy simulations. The feature is a black-box model that requires extensive performance data for each hybrid unitary product. In parallel, we also developed new features for the Technology Performance Exchange to enable manufacturers to submit performance data in a standard format that can be used with the hybrid unitary model in EnergyPlus. Additionally, through this project we expanded university educational resources, and university- manufacturing industry collaborations in the field of energy efficiency technology. Over two years, we involved 20 undergraduate students in ambitious research projects focused on modeling complex multi-mode mechanical systems, supported three mechanical engineering bachelor theses, established undergraduate apprenticeships with multiple industry partners, and involved those partners in the process of design, validation, and debugging for the new EnergyPlus feature. The EnergyPlus feature is described and discussed in an academic article, as well as in an engineering reference, and input/output reference documentation for EnergyPlus. The Technology Performance Exchange features are live and publicly accessible, our manufacturer partners are primed to submit initial product information and performance data to the exchange, and the EnergyPlus feature is scheduled for public release in Spring 2018 as a part of EnergyPlus v8.9.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Scalable Contour Tree Computation by Data Parallel Peak Pruning

As data sets grow to exascale, automated data analysis and visualization are increasingly important, to intermediate human understanding and to reduce demands on disk storage via in situ analysis. Trends in architecture of high performance computing systems necessitate analysis algorithms to make effective use of combinations of massively multicore and distributed systems. One of the principal analytic tools is the contour tree, which analyses relationships between contours to identify features of more than local importance. Unfortunately, the predominant algorithms for computing the contour tree are explicitly serial, and founded on serial metaphors, which has limited the scalability of this form of analysis. While there is some work on distributed contour tree computation, and separately on hybrid GPU-CPU computation, there is no efficient algorithm with strong formal guarantees on performance allied with fast practical performance. Furthermore, we report the first shared SMP algorithm for fully parallel contour tree computation, with formal guarantees of O(lg V lg t) parallel steps and O(V lg V) work for data with V samples and t contour tree supernodes, and implementations with more than 30× parallel speed up on both CPU using TBB and GPU using Thrust and up 70× speed up compared to the serial sweep and merge algorithm.

97 MATHEMATICS AND COMPUTING↗

Energy efficient indoor VOC air cleaning with activated carbon fiber (ACF) filters

This study explores the potential environmental and energy benefits of using activated carbon fiber (ACF) filters for air cleaning in HVAC systems. The parallel aims for the air cleaning system were to enable reduced indoor exposures to volatile organic compounds (VOCs) and to simultaneously allow reduced rates and energy consumption for outdoor-air ventilation. We evaluated the use of ACF media to adsorb VOCs from indoor air during repeated simulated 12-hour to 24-hour periods of occupancy. In a cyclic regeneration process, VOCs were desorbed from the ACF media and vented outdoors to enable the next cycle of air cleaning. The VOC removal efficiency of the ACF media was measured using a 9.5-cm 2 ACF specimen exposed to a mixture of VOCs that included toluene, benzene, o-xylene, 1-butanol, limonene, undecane and formaldehyde at 29 °C and 30 percent relative humidity. The concentrations of these model pollutants upstream of the ACF media were in the range 20 to 30 ppb, to simulate realistic conditions. Velocities through the ACF media were typical of those in normal particle filter systems (~ 0.5 m s -1 ). Initial tests were conducted to develop a modified multi-component Freundlich isotherm and estimate the maximum adsorption capacity of the media, which was determined to be 90 mg VOC per gram of ACF. Three different ACF regeneration methods were explored using relatively cleaner outdoor air under ambient conditions, with this air humidified, and with the filter heated. It was found that heating the ACF media to ~150 °C by circulation of a DC current through the fibers for a short period (15 minutes) yielded the best VOC removal results, allowing for subsequent consistent removal efficiencies of 70-80 percent for most VOCs. Regeneration with unheated outdoor air was also effective and used less energy (subsequent removal efficiency was 50-60 percent for most VOCs). ACF did not perform as well in eliminating formaldehyde, for which a maximum removal of 25-30 percent was achieved with heated regeneration. A mass balance model indicated that the combination of ACF air cleaning and a 50percent reduction in ventilation will decrease indoor concentrations of VOCs by 60 percent to 80 percent and reduce formaldehyde concentrations by 12 percent to 40 percent. Energy modeling indicated the potential to reduce the energy required for heating and cooling of ventilation air by 35 percent to almost 50 percent.

54 ENVIRONMENTAL SCIENCES↗

MapReduceXMT v. Beta 0.1

The MapReduceXMT library ports the MapReduce framework onto the Cray XMT. MapReduce is a programming paradigm and an approach to data management for unstructured problems. It has gained relevance due to its ability to map serial operations onto parallel distributed architectures, significantly improving developer/analyst productivity. The MapReduceXMT implements the key aspects of MapReduce for the Cray XMT, a massively threaded system that is inherently difficult to program. MapReduceXMT allows users to utilize the machine effectively and efficiently without extensive training in multi-threaded programming. The MapReduceXMT library ports the MapReduce framework onto the Cray XMT. MapReduce is a programming paradigm and an approach to data management for unstructured problems. It has gained relevance due to its ability to map serial operations onto parallel distributed architectures, significantly improving developer/analyst productivity. The MapReduceXMT implements the key aspects of MapReduce for the Cray XMT, a massively threaded system that is inherently difficult to program. MapReduceXMT allows users to utilize the machine effectively and efficiently without extensive training in multi-threaded programming.

Goodman, Eric↗

The heat transfer coefficient associated with a moving packed bed of silica particles flowing through parallel plates

Concentrating Solar Power (CSP) with thermal energy storage has the potential to be a renewable energy technology with long duration, inexpensive energy storage. Higher temperature operation increases the conversion efficiency and reduces the cost of energy storage. Several emerging CSP designs utilize particles as the solar receiver due to their high temperature stability and low cost. In some designs, the particles are also used as the thermal energy storage (TES) media. In either configuration, an energy transfer is required between the hot particles and the working fluid in the power cycle in a Particle-to-Fluid Heat Exchanger (PtFHX). Understanding the heat transfer between a moving packed bed and a stationary surface is critical to the successful design of a PtFHX. In this paper, a test facility is described in which a moving packed bed of silica sand with particle size 100–600 μm is introduced into the channel formed by two parallel plates, one of which is heated. The effective static thermal conductivity of the particles used for the test are separately measured over the entire range of test temperatures. The inlet and outlet bulk temperatures of the particle flow are measured as are the surface temperatures at several axial locations along the centerline of the plate. The result is the measurement of heat transfer coefficient as a function of temperature for several velocities. The uncertainty of the measurements is presented and the results are compared to model results found in the literature.

14 SOLAR ENERGY↗

Dynamic model-based sensor network design algorithm for system efficiency maximization

Here, a dynamic model-based sensor network design (DMSND) algorithm has been developed for maximizing system efficiency for an estimator-based control system. The algorithm synthesizes the optimal sensor network in the face of disturbances or set point changes. Computational expense of the large-scale combinatorial optimization problem is significantly reduced by parallel computing and by using combination of three novel strategies: multi-rate sampling frequency, model order reduction, and use of an incumbent solution that enables early termination of evaluation of infeasible sensor sets. The developed algorithm is applied to an acid gas removal unit as part of an integrated gasification combined cycle power plant with carbon capture. Even though there are more than thousand process states and more than hundred candidate sensor locations, the optimal sensor network design problem for maximizing process efficiency could be solved within couple of hours for a given budget.

47 OTHER INSTRUMENTATION↗

A Comparison of Shared Memory Parallel Programming Models

The dominant parallel programming models for shared memory computers, Pthreads and OpenMP, are both thread-centric in that they are based on explicit management of tasks and enforce data dependencies and output ordering through task management. By comparison, the Cray XMT programming model is data-centric where the primary concern of the programmer is managing data dependencies, allowing threads to progress in a data flow fashion. The XMT implements this programming model by associating tag bits with each word of memory, affording efficient fine grained data synchronization independent of the number of processors or how tasks are scheduled. When task management is implicit and synchronization is abundant, efficient, and easy to use, programmers have viable alternatives to traditional thread-centric algorithms. In this paper we compare the amount of available parallelism relative to the amount of work in a variety of different algorithms and data structures when synchronization does not need to be rationed, as well as identify opportunities for platform and performance portability of the data-centric programming model on multi-core processors.

99 GENERAL AND MISCELLANEOUS↗

A Comparison of Two Yeast MnSODs: Mitochondrial Saccharomyces cerevisiae versus Cytosolic Candida albicans

Human MnSOD is significantly more product-inhibited than bacterial MnSODs at high concentrations of superoxide (O{sub 2}{sup -}). This behavior limits the amount of H{sub 2}O{sub 2} produced at high [O{sub 2}{sup -}]; its desirability can be explained by the multiple roles of H{sub 2}O{sub 2} in mammalian cells, particularly its role in signaling. To investigate the mechanism of product inhibition in MnSOD, two yeast MnSODs, one from Saccharomyces cerevisiae mitochondria (ScMnSOD) and the other from Candida albicans cytosol (CaMnSODc), were isolated and characterized. ScMnSOD and CaMnSODc are similar in catalytic kinetics, spectroscopy, and redox chemistry, and they both rest predominantly in the reduced state (unlike most other MnSODs). At high [O{sub 2}{sup -}], the dismutation efficiencies of the yeast MnSODs surpass those of human and bacterial MnSODs, due to very low level of product inhibition. Optical and parallel-mode electron paramagnetic resonance (EPR) spectra suggest the presence of two Mn{sup 3+} species in yeast Mn{sup 3+}SODs, including the well-characterized 5-coordinate Mn{sup 3+} species and a 6-coordinate L-Mn{sup 3+} species with hydroxide as the putative sixth ligand (L). The first and second coordination spheres of ScMnSOD are more similar to bacterial than to human MnSOD. Gln154, an H-bond donor to the Mn-coordinated solvent molecule, is slightly further away from Mn in yeast MnSODs, which may result in their unusual resting state. Mechanistically, the high efficiency of yeast MnSODs could be ascribed to putative translocation of an outer-sphere solvent molecule, which could destabilize the inhibited complex and enhance proton transfer from protein to peroxide. Our studies on yeast MnSODs indicate the unique nature of human MnSOD in that it predominantly undergoes the inhibited pathway at high [O{sub 2}{sup -}].

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗

HARD: A performance portable radiation hydrodynamics code based on FleCSI framework

Hydrodynamics And Radiation Diffusion (HARD) is an open-source application for high-performance simulations of compressible hydrodynamics with radiation-diffusion coupling. Built on the FleCSI (Bergen et al., 2021 [1]) (Flexible Computational Science Infrastructure) framework, HARD expresses its computational units as tasks whose execution can be orchestrated by multiple back-end runtimes, including Legion (Bauer et al., 2012 [2]), MPI (Forum, 1994 [3]), and HPX (Kaiser et al., 2020 [4]). Node-level parallelism is handled through Kokkos (Edwards et al., 2014 [5]), providing a single-source, portable code base that runs efficiently on laptops, small homogeneous clusters, and the largest heterogeneous supercomputers currently available. To ensure scientific reliability, HARD includes a regression test suite that automatically reproduces canonical verification problems such as the Sod and LeBlanc shock tubes, and the Sedov blast wave, comparing numerical solutions against known analytical results. The project is distributed under an OSI-approved license, hosted on GitHub, and accompanied by reproducible build scripts and continuous integration workflows. This combination of performance portability, verification infrastructure, and community-focused development makes HARD a sustainable platform for advancing radiation hydrodynamics research across multiple domains.

97 MATHEMATICS AND COMPUTING↗

Performance analysis of distributed symmetric sparse matrix vector multiplication algorithm for multi-core architectures

In this article, sparse matrix vector multiply (SpMVM) is an important kernel that frequently arises in high performance computing applications. Due to its low arithmetic intensity, several approaches have been proposed in literature to improve its scalability and efficiency in large scale computations. In this paper, our target systems are high end multi-core architectures and we use messaging passing interface + open multiprocessing hybrid programming model for parallelism. We analyze the performance of recently proposed implementation of the distributed symmetric SpMVM, originally developed for large sparse symmetric matrices arising in ab initio nuclear structure calculations. We also study important features of this implementation and compare with previously reported implementations that do not exploit underlying symmetry. Our SpMVM implementations leverage the hybrid paradigm to efficiently overlap expensive communications with computations. Our main comparison criterion is the "CPU core hours" metric, which is the main measure of resource usage on supercomputers. We analyze the effects of topology-aware mapping heuristic using simplified network load model. Furthermore, we have tested the different SpMVM implementations on two large clusters with 3D Torus and Dragonfly topology. Our results show that the distributed SpMVM implementation that exploits matrix symmetry and hides communication yields the best value for the "CPU core hours" metric and significantly reduces data movement overheads.

97 MATHEMATICS AND COMPUTING↗

Tethys: Developing a Commons for Understanding Environmental Effects of Ocean Renewable Energy

A lack of consistent widespread information on potential environmental effects, monitoring outcomes, and mitigation strategies continues to hinder the development and deployment of marine energy systems around the world. To address this need, member nations of the Ocean Energy Systems (OES) formed Annex IV, a three-year collaborative effort (2010-2013) to increase knowledge sharing, research collaborations, and efficient regulatory decision-making associated with investigating and evaluating the potential impacts of wave and tidal energy. In parallel, the US Department of Energy began development of a publically-available, searchable online database of environmental information (called Tethys), housing scientific literature pertaining to the environmental effects of marine energy and offshore wind systems, as well as metadata on international ocean energy projects and research studies collected under Annex IV. In addition to providing broad access to information on environmental effects of marine energy and offshore wind developments, Tethys can act as a commons for bringing together researchers, device and project developers, regulators, and stakeholders, to explore shared interests and information needs for the emerging marine energy industry. This paper will demonstrate the development, existing functionality, and content of Tethys, as well as discuss the future of Annex IV including information updates, improvements, and additional outreach efforts over the next several years.

marine energy, offhore wind, environmental effects↗

Rapid measurement of sub-micrometer aerosol size distribution using a fast integrated mobility spectrometer

We present rapid measurement of submicron particle size distributions enables the characterization of aerosols with fast changing properties, and is often necessary for measurements onboard mobile platforms (e.g., research aircraft). Aerosol mobility size distribution is commonly measured by a scanning mobility particle sizer (SMPS), which relies on voltage scanning or stepping to classify particles of different sizes, and may take up to several minutes to obtain a complete size spectrum of aerosol particles. The recently developed fast integrated mobility spectrometer (FIMS) with enhanced dynamic size range classifies and detects particles from 10 to ~600 nm simultaneously, allowing submicron aerosol mobility size distributions to be captured at a time resolution of 1 second. In this study, we present a detailed data inversion routine for deriving aerosol size distribution from FIMS measurements. The inversion routine takes into consideration the FIMS transfer function, particle penetration efficiency in the FIMS, and multiple charging of aerosols. The accuracy of the FIMS measurement is demonstrated by comparing parallel FIMS and SMPS measurements of stable aerosols with a wide range of size spectrum shapes, including ambient aerosols and aerosols classified by a differential mobility analyzer (DMA). The FIMS and SMPS-derived size distributions show excellent agreements for all aerosols tested. In addition, total number concentrations of ambient aerosols were integrated from 1 Hz FIMS size distributions, and compared with those directly measured by a condensation particle counter (CPC) operated in parallel. Finally, the integrated and measured total particle concentrations agree well within 5%.

54 ENVIRONMENTAL SCIENCES↗

Optimizing the hypre solver for manycore and GPU architectures

The solution of large-scale combustion problems with codes such as Uintah on modern computer architectures requires the use of multithreading and GPUs to achieve performance. Uintah uses a low-Mach number approximation that requires iteratively solving a large system of linear equations. The Hypre iterative solver has solved such systems in a scalable way for Uintah, but the use of OpenMP with Hypre leads to at least slowdown due to OpenMP overheads. The proposed solution uses the MPI Endpoints within Hypre, where each team of threads acts as a different MPI rank. This approach minimizes OpenMP synchronization overhead and performs as fast or (up to 1.44) faster than Hypre's MPI-only version, and allows the rest of Uintah to be optimized using OpenMP. The profiling of the GPU version of Hypre shows the bottleneck to be the launch overhead of thousands of micro-kernels. The GPU performance was improved by fusing these micro-kernels and was further optimized by using Cuda-aware MPI, resulting in an overall speedup of 1.16—1.44 compared to the baseline GPU implementation. The above optimization strategies were published in the International Conference on Computational Science 2020 [1]. This work extends the previously published research by carrying out the second phase of communication-centered optimizations in Hypre to improve its scalability on large-scale supercomputers. Additionally, this includes an efficient non-blocking inter-thread communication scheme, communication-reducing patch assignment, and expression of logical communication parallelism to a new version of the MPICH library that utilizes the underlying network parallelism [2]. The above optimizations avoid communication bottlenecks previously observed during strong scaling and improve performance by up to 2 on 256 nodes of Intel Knight's Landing processor.

97 MATHEMATICS AND COMPUTING↗

Phosphonium-Containing Block Copolymer Anion Exchange Membranes: Effect of Quaternization Level on Bulk and Surface Morphologies at Hydrated and Dehydrated States

Diblock copolymers of isoprene and chloromethylstyrene were synthesized by nitroxide-mediate polymerization. The polychloromethylstyrene block was quaternized with tris(trimethoxyphenyl)phosphine (P(Ph(OMe) 3 ) 3 ) or triphenyl-phosphine (P(Ph) 3 ). Membranes with ion-exchange capacity (IEC) ranging from 0.87 to 2.35 mmol/g were prepared. The bulk morphology was investigated via small-angle X-ray scattering and transmission electron microscopy, where hexagonal and body-centered cubic morphologies were observed. The surface morphology was revealed by atomic force microscopy where both parallel and perpendicular alignments were observed. Compared to P(Ph(OMe) 3 ) 3 , P(Ph) 3 exhibits higher quaternization efficiency due to less steric hindrance. Partially quaternized membranes were prepared using P(Ph) 3 , and the dependence of domain size on the quaternization level is discussed. The morphologies as a function of relative humidity were investigated. For high IEC membranes, although bulk scattering showed small changes in d -spacing, large changes in domain size on the surface occurred between hydrated and dehydrated conditions. Finally, further investigations of morphological reversibility by humidity cycling showed good recovery between humid and dry conditions, indicating that the expansion and contraction of channels is a reversible process.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Evanescent Mode Photoemission

Photoemission of electrons from solid surfaces into vacuum is routinely used in two configurations: reflection mode, in which light is incident on the emitting surface, and transmission mode, in which light illuminates the photoemissive material from behind. Here, using silicon nitride waveguides integrated beneath a high quantum efficiency cesium antimonide thin film, we demonstrate a new photoemission regime─the evanescent mode photoemission─in which light travels parallel to the surface in the waveguide while evanescently coupling into the thin film to excite and emit photoelectrons. Furthermore, our experiments show that this configuration enables direct visualization of guided optical modes over an unprecedented range (∼100 nm to 1 mm) and allows electron beam shaping at the source with transverse features well below ∼600 nm, establishing evanescent mode photoemission as a new platform for nanoscale optical field mapping and precise generation of electron beams.

Electron Sources↗