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 1,225 records · Page 68

Driving Down HB-LED Costs. Implementation of Process Simulation Tools and Temperature Control Methods of High Yield MOCVD Growth

The overall objective of this multi-faceted program is to develop epitaxial growth systems that meet a goal of 75% (4X) cost reduction in the epitaxy phase of HB-LED manufacture. A 75% reduction in yielded epitaxy cost is necessary in order to achieve the cost goals for widespread penetration of HB-LED's into back-lighting units (BLU) for LCD panels and ultimately for solid-state lighting (SSL). To do this, the program will address significant improvements in overall equipment Cost of Ownership, or CoO. CoO is a model that includes all costs associated with the epitaxy portion of production. These aspects include cost of yield, capital cost, operational costs, and maintenance costs. We divide the program into three phases where later phases will incorporate the gains of prior phases. Phase one activities are enabling technologies. In collaboration with Sandia National Laboratories we develop a Fluent-compatible chemistry predictive model and a set of mid-infrared and near-ultraviolet pyrometer monitoring tools. Where previously the modeling of the reactor dynamics were studied within FLUENT alone, here, FLUENT and Chemkin are integrated into a comprehensive model of fluid dynamics and the most advanced transport equations developed for Chemkin. Specifically, the Chemkin model offered the key reaction terms for gas-phase nucleation, a key consideration in the optimization of the MOCVD process. This new predictive model is used to design new MOCVD reactors with optimized growth conditions and the newly developed pyrometers are used monitor and control the MOCVD process temperature to within 0.5°C run-to-run and within each wafer. This portion of the grant is in collaboration with partners at Sandia National Laboratories. Phase two activities are continuous improvement projects which extend the current reactor platform along the lines of improved operational efficiency, improved systems control for throughput, and carrier modifications for increased yield. Programmatically, improvements made in Phase I are applied to developments of Phase II when applicable. Phase three is the culmination of the individual tasks from both phases one and two applied to proposed production platforms. We selectively combine previously demonstrated tasks and other options to develop a high-volume production-worthy MOCVD system demonstrating >3x throughput, 1.3x capital efficiency, and 0.7x cost of ownership. In a parallel demonstration we validate the concept of an improved, larger deposition system which utilizes the predictive modeling of chemistry-based flow analysis and extensions of the improvements demonstrated on the current platforms. This validation includes the build and testing of a prototype version of the hardware and demonstration of 69% reduction in the cost of ownership. Also, in this phase we present a stand-alone project to develop a high-temperature system which improves source efficiency by 30% while concurrently increasing growth rate by 1.3x. The material quality is held to the same material quality specifications of our existing baseline processes. The merits of other line item tasks in phase three are discussed for inclusion on next-generation platforms.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Parallel quantum annealing

Quantum annealers of D-Wave Systems, Inc., offer an efficient way to compute high quality solutions of NP-hard problems. This is done by mapping a problem onto the physical qubits of the quantum chip, from which a solution is obtained after quantum annealing. However, since the connectivity of the physical qubits on the chip is limited, a minor embedding of the problem structure onto the chip is required. In this process, and especially for smaller problems, many qubits will stay unused. We propose a novel method, called parallel quantum annealing, to make better use of available qubits, wherein either the same or several independent problems are solved in the same annealing cycle of a quantum annealer, assuming enough physical qubits are available to embed more than one problem. Although the individual solution quality may be slightly decreased when solving several problems in parallel (as opposed to solving each problem separately), we demonstrate that our method may give dramatic speed-ups in terms of the Time-To-Solution (TTS) metric for solving instances of the Maximum Clique problem when compared to solving each problem sequentially on the quantum annealer. Additionally, we show that solving a single Maximum Clique problem using parallel quantum annealing reduces the TTS significantly.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulated evolution of fluorophores for light emitting diodes

Organic light emitting diodes based on fluorophores with a propensity for thermally activated delayed fluorescence (TADF) are able to circumvent limitations imposed on device efficiency by spin statistics. Molecules with a propensity for TADF necessarily have two properties: a small gap between the lowest lying singlet and triplet excited states and a large transition dipole moment for fluorescence. In this work, we demonstrate the use of a genetic algorithm to search a region of chemical space for molecules with these properties. This algorithm is based on a flexible and intuitive representation of the molecule as a tree data structure, in which the nodes correspond to molecular fragments. Our implementation takes advantage of hybrid parallel graphics processing unit accelerated computer clusters to allow efficient sampling while retaining a reasonably accurate description of the electronic structure (in this case, CAM-B3LYP/6-31G{sup ∗∗}). In total, we have identified 3792 promising candidate fluorophores from a chemical space containing 1.26 × 10{sup 6} molecules. This required performing electronic structure calculations on only 7518 molecules, a small fraction of the full space. Several novel classes of molecules which show promise as fluorophores are presented.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DeepClone: Lightweight State Replication of Deep Learning Models for Data Parallel Training

Training modern deep neural network (DNN) models involves complex workflows triggered by model exploration, sensitivity analysis, explainability, etc. A key primitive in this context is the ability to clone a model training instance, i.e. “fork” the training process in a potentially different direction, which enables comparisons of different evolution paths using variations of training data and model parameters. However, in a quest improve the training throughput, a mix of data parallel, model parallel, pipeline parallel and layer-wise parallel approaches are making the problem of cloning highly complex. In this paper, we explore the problem of efficient cloning under such circumstances. To this end, we leverage several properties of data-parallel training and layer-wise parallelism to design DeepClone, a cloning approach based on augmenting the execution graph to gain direct access to tensors, which are then sharded and reconstructed asynchronously in order to minimize runtime overhead, standby duration, readiness duration. Compared with state-of-art approaches, DeepClone shows orders of magnitude improvement for several classes of DNN models.

data-parallel training↗

Parallel computing for probabilistic fatigue analysis

This paper presents the results of Phase I research to investigate the most effective parallel processing software strategies and hardware configurations for probabilistic structural analysis. We investigate the efficiency of both shared and distributed-memory architectures via a probabilistic fatigue life analysis problem. We also present a parallel programming approach, the virtual shared-memory paradigm, that is applicable across both types of hardware. Using this approach, problems can be solved on a variety of parallel configurations, including networks of single or multiprocessor workstations. We conclude that it is possible to effectively parallelize probabilistic fatigue analysis codes; however, special strategies will be needed to achieve large-scale parallelism to keep large number of processors busy and to treat problems with the large memory requirements encountered in practice. We also conclude that distributed-memory architecture is preferable to shared-memory for achieving large scale parallelism; however, in the future, the currently emerging hybrid-memory architectures will likely be optimal.

Sues, Robert H.↗

Impedance Matching a Quarter Wave Resonant Receiver to Improve Efficiency in Unipolar Capacitive Wireless Power Transfer

Impedance matching techniques to achieve high performance in conventional wireless power transfer (WPT) systems are well established. These WPT systems are modelled as a lumped circuit, where the voltages/currents are approximately uniform and mostly unaffected by parasitic elements. However, a group of WPT techniques exist that utilize parasitic elements to form a highly resonant power transfer condition at high quality factors, exploiting a distributed circuit operation that can be modeled as a mutually coupled transmission line. A calculated matching network typically becomes unmatched when connected to the transmission line, as it changes the electrical structure of the coupled transmission line. This is particularly true for WPT systems that use quarter wave (QWR) or half wave resonators (HWR) for the receiver. Here we present one technique of matching to a mutually coupled transmission line where a parallel path is formed that improves the power transfer efficiency without severely altering the transmission line characteristics.

Wireless↗

Analyzing Tropical Waves Using the Parallel Ensemble Empirical Model Decomposition Method: Preliminary Results from Hurricane Sandy

In this study, we discuss the performance of the parallel ensemble empirical mode decomposition (EMD) in the analysis of tropical waves that are associated with tropical cyclone (TC) formation. To efficiently analyze high-resolution, global, multiple-dimensional data sets, we first implement multilevel parallelism into the ensemble EMD (EEMD) and obtain a parallel speedup of 720 using 200 eight-core processors. We then apply the parallel EEMD (PEEMD) to extract the intrinsic mode functions (IMFs) from preselected data sets that represent (1) idealized tropical waves and (2) large-scale environmental flows associated with Hurricane Sandy (2012). Results indicate that the PEEMD is efficient and effective in revealing the major wave characteristics of the data, such as wavelengths and periods, by sifting out the dominant (wave) components. This approach has a potential for hurricane climate study by examining the statistical relationship between tropical waves and TC formation.

PEEMD↗

Efficient Ensemble-Based Stochastic Gradient Methods for Optimization Under Geological Uncertainty

Ensemble-based stochastic gradient methods, such as the ensemble optimization (EnOpt) method, the simplex gradient (SG) method, and the stochastic simplex approximate gradient (StoSAG) method, approximate the gradient of an objective function using an ensemble of perturbed control vectors. These methods are increasingly used in solving reservoir optimization problems because they are not only easy to parallelize and couple with any simulator but also computationally more efficient than the conventional finite-difference method for gradient calculations. In this work, we show that EnOpt may fail to achieve sufficient improvement of the objective function when the differences between the objective function values of perturbed control variables and their ensemble mean are large. On the basis of the comparison of EnOpt and SG, we propose a hybrid gradient of EnOpt and SG to save on the computational cost of SG. We also suggest practical ways to reduce the computational cost of EnOpt and StoSAG by approximating the objective function values of unperturbed control variables using the values of perturbed ones. We first demonstrate the performance of our improved ensemble schemes using a benchmark problem. Results show that the proposed gradients saved about 30–50% of the computational cost of the same optimization by using EnOpt, SG, and StoSAG. As a real application, we consider pressure management in carbon storage reservoirs, for which brine extraction wells need to be optimally placed to reduce reservoir pressure buildup while maximizing the net present value. Results show that our improved schemes reduce the computational cost significantly.

58 GEOSCIENCES↗

SU-E-T-499: Initial Developments of An OpenCL-Based Cross-Platform Monte Carlo Dose Engine for Carbon Ion Therapy

Purpose Dose calculation is of critical importance for carbon ion therapy. Monte Carlo (MC) simulation is considered to be the most accurate method for calculation of absorbed dose and of all the more fundamental physical quantities related to biological effects. The long computation time, however, limits its routine clinical applications. We have recently started developing a fast MC package, gCMC for carbon therapy on a parallel processing platform, e.g. GPU, aiming at achieving sufficient efficiency to enable MC in clinically important tasks. This abstract reports our progress. Methods gCMC was developed in OpenCL environment. Our initial developments focused on water material. gCMC supported carbon ion transport in the energy range of 1–450 MeV/u. A Class II condensed history algorithm was implemented for charged particle transport simulations with stopping power computed via Bethe-Bloch equation. Energy straggling and multiple scattering were modeled. Total cross section of nuclear interaction was extracted from Geant4. At present, nuclear interaction events were sampled but transports of secondary particles were not included. Results We tested cases with a homogeneous water phantom and a pencil carbon ion beam with energy of 200–400 MeV/u. When only electro-magnetic channel was included, dose/fluence difference between gCMC and Geant4 results averaged within 10% isodose line was <0.5% of the maximum dose/fluence. After enabling nuclear interactions without transporting secondary particles, dose and fluence agreed with the corresponding results computed by Geant4 with <1% difference. Due to the support for multiple platforms of OpenCL, gCMC was executable on NVidia and AMD GPUs, and Intel CPUs. It took ∼50 sec to transport 107 200MeV/u source carbon ions on an NVidia Titan GPU card. Conclusion Preliminary studies have demonstrated the accuracy and efficiency of gCMC. With further developments in near future, gCMC will potentially achieve clinically acceptable fast and accurate MC simulations for carbon ion therapy.

60 APPLIED LIFE SCIENCES↗

Multipinhole collimator with 20 apertures for a brain SPECT application

Purpose: Several new technologies for single photon emission computed tomography (SPECT) instrumentation with parallel-hole collimation have been proposed to improve detector sensitivity and signal collection efficiency. Benefits from improved signal efficiency include shorter acquisition times and lower dose requirements. In this paper, the authors show a possibility of over an order of magnitude enhancement in photon detection efficiency (from 7.6 × 10{sup −5} to 1.6 × 10{sup −3}) for dopamine transporter (DaT) imaging of the striatum over the conventional SPECT parallel-hole collimators by use of custom-designed 20 multipinhole (20-MPH) collimators with apertures of 0.75 cm diameter. Methods: Quantifying specific binding ratio (SBR) of {sup 123}I-ioflupane or {sup 123}I-iometopane’s signal at the striatal region is a common brain imaging method to confirm the diagnosis of the Parkinson’s disease. The authors performed imaging of a striatal phantom filled with aqueous solution of I-123 and compared camera recovery ratios of SBR acquired between low-energy high-resolution (LEHR) parallel-hole collimators and 20-MPH collimators. Results: With only two-thirds of total acquisition time (20 min against 30 min), a comparable camera recovery ratio of SBR was achieved using 20-MPH collimators in comparison to that from the LEHR collimator study. Conclusions: Their systematic analyses showed that the 20-MPH collimator could be a promising alternative for the DaT SPECT imaging for brain over the traditional LEHR collimator, which could give both shorter scan time and improved diagnostic accuracy.

60 APPLIED LIFE SCIENCES↗

Data Movement Dominates: Advanced Memory Technology to Address the Real Exascale Power Problem

Energy is the fundamental barrier to Exascale supercomputing and is dominated by the cost of moving data from one point to another, not computation. Similarly, performance is dominated by data movement, not computation. The solution to this problem requires three critical technologies: 3D integration, optical chip-to-chip communication, and a new communication model. The central goal of the Sandia led "Data Movement Dominates" project aimed to develop memory systems and new architectures based on these technologies that have the potential to lower the cost of local memory accesses by orders of magnitude and provide substantially more bandwidth. Only through these transformational advances can future systems reach the goals of Exascale computing with a manageable power budgets. The Sandia led team included co-PIs from Columbia University, Lawrence Berkeley Lab, and the University of Maryland. The Columbia effort of Data Movement Dominates focused on developing a physically accurate simulation environment and experimental verification for optically-connected memory (OCM) systems that can enable continued performance scaling through high-bandwidth capacity, energy-efficient bit-rate transparency, and time-of-flight latency. With OCM, memory device parallelism and total capacity can scale to match future high-performance computing requirements without sacrificing data-movement efficiency. When we consider systems with integrated photonics, links to memory can be seamlessly integrated with the interconnection network-in a sense, memory becomes a primary aspect of the interconnection network. At the core of the Columbia effort, toward expanding our understanding of OCM enabled computing we have created an integrated modeling and simulation environment that uniquely integrates the physical behavior of the optical layer. The PhoenxSim suite of design and software tools developed under this effort has enabled the co-design of and performance evaluation photonics-enabled OCM architectures on Exascale computing systems.

97 MATHEMATICS AND COMPUTING↗

A Robust and Scalable Software Library for Parallel Adaptive Refinement on Unstructured Meshes

The design and implementation of Pyramid, a software library for performing parallel adaptive mesh refinement (PAMR) on unstructured meshes, is described. This software library can be easily used in a variety of unstructured parallel computational applications, including parallel finite element, parallel finite volume, and parallel visualization applications using triangular or tetrahedral meshes. The library contains a suite of well-designed and efficiently implemented modules that perform operations in a typical PAMR process. Among these are mesh quality control during successive parallel adaptive refinement (typically guided by a local-error estimator), parallel load-balancing, and parallel mesh partitioning using the ParMeTiS partitioner. The Pyramid library is implemented in Fortran 90 with an interface to the Message-Passing Interface (MPI) library, supporting code efficiency, modularity, and portability. An EM waveguide filter application, adaptively refined using the Pyramid library, is illustrated.

Lou, John Z.↗

A Parallel Computing Infrastructure for Building Energy Simulation

In order to study grid-interactive efficient buildings, Pacific Northwest National Laboratories (PNNL) needs an infrastructure for urban-scale building energy modeling. Such an infrastructure should be fast, scalable, and easy-to-use. Given a set of data from the Energy Information Administration’s Commercial Building Energy Consumption Survey (CBECS) and tool to translate survey data into simulation inputs, this project aimed to conduct the simulation of the entire dataset in parallel. Before running the simulations, the necessary software was bundled into a container for use on the PNNL supercomputing network. Then, the parallel simulation workflow was designed using GNU Make, a file creation software, and submitted to a supercomputing partition which could run hundreds of simulations simultaneously. The EnergyPlus simulations output hourly electric meter data for each CBECS sample, which represents the electricity consumption of similar commercial buildings across the United States. Analyzing and visualizing the meter data is important to the future of the work, and this project wrote code to make common analysis methods simple, fast, and accessible. Moving forwards, the model will need to be expanded to include data from other sources and its accuracy will need to be improved and eventually validated.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Statistical and machine learning models for optimizing energy in parallel applications

Rising power costs and constraints are driving a growing focus on the energy efficiency of high performance computing systems. The unique characteristics of a particular system and workload and their effect on performance and energy efficiency are typically difficult for application users to assess and to control. Settings for optimum performance and energy efficiency can also diverge, so we need to identify trade-off options that guide a suitable balance between energy use and performance. We present statistical and machine learning models that only require a small number of runs to make accurate Pareto-optimal trade-off predictions using parameters that users can control. We study model training and validation using several parallel kernels and more complex workloads, including Algebraic Multigrid (AMG), Large-scale Atomic Molecular Massively Parallel Simulator, and Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. Here, we demonstrate that we can train the models using as few as 12 runs, with prediction error of less than 10%. Our AMG results identify trade-off options that provide up to 45% improvement in energy efficiency for around 10% performance loss. We reduce the sample measurement time required for AMG by 90%, from 13 h to 74 min.

97 MATHEMATICS AND COMPUTING↗

Fast correlation function calculator: A high-performance pair-counting toolkit

A novel high-performance exact pair-counting toolkit called fast correlation function calculator (FCFC) is presented. With the rapid growth of modern cosmological datasets, the evaluation of correlation functions with observational and simulation catalogues has become a challenge. High-efficiency pair-counting codes are thus in great demand. We introduce different data structures and algorithms that can be used for pair-counting problems, and perform comprehensive benchmarks to identify the most efficient algorithms for real-world cosmological applications. We then describe the three levels of parallelisms used by FCFC, SIMD, OpenMP, and MPI, and run extensive tests to investigate the scalabilities. Finally, we compare the efficiency of FCFC with alternative pair-counting codes. The data structures and histogram update algorithms implemented in FCFC are shown to outperform alternative methods. FCFC does not benefit greatly from SIMD because the bottleneck of our histogram update algorithm is mainly cache latency. Nevertheless, the efficiency of FCFC scales well with the numbers of OpenMP threads and MPI processes, even though speedups may be degraded with over a few thousand threads in total. FCFC is found to be faster than most (if not all) other public pair-counting codes for modern cosmological pair-counting applications.

79 ASTRONOMY AND ASTROPHYSICS↗

An adaptive discontinuous Petrov-Galerkin method for the Grad-Shafranov equation

In this work, we propose and develop an arbitrary-order adaptive discontinuous Petrov--Galerkin (DPG) method for the nonlinear Grad--Shafranov equation. An ultraweak formulation of the DPG scheme for the equation is given based on a minimal residual method. The DPG scheme has the advantage of providing more accurate gradients compared to conventional finite element methods, which is desired for numerical solutions to the Grad--Shafranov equation. The numerical scheme is augmented with an adaptive mesh refinement approach, and a criterion based on the residual norm in the minimal residual method is developed to achieve dynamic refinement. Nonlinear solvers for the resulting system are explored and a Picard iteration with Anderson acceleration is found to be efficient to solve the system. Finally, the proposed algorithm is implemented in parallel on MFEM using a domain-decomposition approach, and our implementation is general, supporting arbitrary order of accuracy and general meshes. Furthermore, numerical results are presented to demonstrate the efficiency and accuracy of the proposed algorithm.

97 MATHEMATICS AND COMPUTING↗

The Computational Complexity, Parallel Scalability, and Performance of Atmospheric Data Assimilation Algorithms

The computational complexity of algorithms for Four Dimensional Data Assimilation (4DDA) at NASA's Data Assimilation Office (DAO) is discussed. In 4DDA, observations are assimilated with the output of a dynamical model to generate best-estimates of the states of the system. It is thus a mapping problem, whereby scattered observations are converted into regular accurate maps of wind, temperature, moisture and other variables. The DAO is developing and using 4DDA algorithms that provide these datasets, or analyses, in support of Earth System Science research. Two large-scale algorithms are discussed. The first approach, the Goddard Earth Observing System Data Assimilation System (GEOS DAS), uses an atmospheric general circulation model (GCM) and an observation-space based analysis system, the Physical-space Statistical Analysis System (PSAS). GEOS DAS is very similar to global meteorological weather forecasting data assimilation systems, but is used at NASA for climate research. Systems of this size typically run at between 1 and 20 gigaflop/s. The second approach, the Kalman filter, uses a more consistent algorithm to determine the forecast error covariance matrix than does GEOS DAS. For atmospheric assimilation, the gridded dynamical fields typically have More than 10(exp 6) variables, therefore the full error covariance matrix may be in excess of a teraword. For the Kalman filter this problem can easily scale to petaflop/s proportions. We discuss the computational complexity of GEOS DAS and our implementation of the Kalman filter. We also discuss and quantify some of the technical issues and limitations in developing efficient, in terms of wall clock time, and scalable parallel implementations of the algorithms.

Lyster, Peter M.↗