Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel performance”

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 1,153 records · Page 64
All sourcesNASA NTRSDOE OSTI
NASA NTRS1989

A Computational Fluid Dynamics Algorithm on a Massively Parallel Computer

The discipline of computational fluid dynamics is demanding ever-increasing computational power to deal with complex fluid flow problems. We investigate the performance of a finite-difference computational fluid dynamics algorithm on a massively parallel computer, the Connection Machine. Of special interest is an implicit time-stepping algorithm; to obtain maximum performance from the Connection Machine, it is necessary to use a nonstandard algorithm to solve the linear systems that arise in the implicit algorithm. We find that the Connection Machine ran achieve very high computation rates on both explicit and implicit algorithms. The performance of the Connection Machine puts it in the same class as today's most powerful conventional supercomputers.

Jespersen, Dennis C.↗
NASA NTRS2000

Incremental Parallelization of Non-Data-Parallel Programs Using the Charon Message-Passing Library

Message passing is among the most popular techniques for parallelizing scientific programs on distributed-memory architectures. The reasons for its success are wide availability (MPI), efficiency, and full tuning control provided to the programmer. A major drawback, however, is that incremental parallelization, as offered by compiler directives, is not generally possible, because all data structures have to be changed throughout the program simultaneously. Charon remedies this situation through mappings between distributed and non-distributed data. It allows breaking up the parallelization into small steps, guaranteeing correctness at every stage. Several tools are available to help convert legacy codes into high-performance message-passing programs. They usually target data-parallel applications, whose loops carrying most of the work can be distributed among all processors without much dependency analysis. Others do a full dependency analysis and then convert the code virtually automatically. Even more toolkits are available that aid construction from scratch of message passing programs. None, however, allows piecemeal translation of codes with complex data dependencies (i.e. non-data-parallel programs) into message passing codes. The Charon library (available in both C and Fortran) provides incremental parallelization capabilities by linking legacy code arrays with distributed arrays. During the conversion process, non-distributed and distributed arrays exist side by side, and simple mapping functions allow the programmer to switch between the two in any location in the program. Charon also provides wrapper functions that leave the structure of the legacy code intact, but that allow execution on truly distributed data. Finally, the library provides a rich set of communication functions that support virtually all patterns of remote data demands in realistic structured grid scientific programs, including transposition, nearest-neighbor communication, pipelining, gather/scatter, and redistribution. At the end of the conversion process most intermediate Charon function calls will have been removed, the non-distributed arrays will have been deleted, and virtually the only remaining Charon functions calls are the high-level, highly optimized communications. Distribution of the data is under complete control of the programmer, although a wide range of useful distributions is easily available through predefined functions. A crucial aspect of the library is that it does not allocate space for distributed arrays, but accepts programmer-specified memory. This has two major consequences. First, codes parallelized using Charon do not suffer from encapsulation; user data is always directly accessible. This provides high efficiency, and also retains the possibility of using message passing directly for highly irregular communications. Second, non-distributed arrays can be interpreted as (trivial) distributions in the Charon sense, which allows them to be mapped to truly distributed arrays, and vice versa. This is the mechanism that enables incremental parallelization. In this paper we provide a brief introduction of the library and then focus on the actual steps in the parallelization process, using some representative examples from, among others, the NAS Parallel Benchmarks. We show how a complicated two-dimensional pipeline-the prototypical non-data-parallel algorithm- can be constructed with ease. To demonstrate the flexibility of the library, we give examples of the stepwise, efficient parallel implementation of nonlocal boundary conditions common in aircraft simulations, as well as the construction of the sequence of grids required for multigrid.

VanderWijngaart, Rob F.↗
NASA NTRS1996

Parallel implementation of an adaptive scheme for 3D unstructured grids on the SP2

Dynamic mesh adaption on unstructured grids is a powerful tool for computing unsteady flows that require local grid modifications to efficiently resolve solution features. For this work, we consider an edge-based adaption scheme that has shown good single-processor performance on the C90. We report on our experience parallelizing this code for the SP2. Results show a 47.0X speedup on 64 processors when 10 percent of the mesh is randomly refined. Performance deteriorates to 7.7X when the same number of edges are refined in a highly-localized region. This is because almost all the mesh adaption is confined to a single processor. However, this problem can be remedied by repartitioning the mesh immediately after targeting edges for refinement but before the actual adaption takes place. With this change, the speedup improves dramatically to 43.6X.

Strawn, Roger C.↗
DOE OSTI2024

Efficient Distributed Sequence Parallelism for Transformer-Based Image Segmentation

We introduce an efficient distributed sequence parallel approach for training transformer-based deep learning image segmentation models. The neural network models are comprised of a combination of a Vision Transformer encoder with a convolutional decoder to provide image segmentation mappings. The utility of the distributed sequence parallel approach is especially useful in cases where the tokenized embedding representation of image data are too large to fit into standard computing hardware memory. To demonstrate the performance and characteristics of our models trained in sequence parallel fashion compared to standard models, we evaluate our approach using a 3D MRI brain tumor segmentation dataset. We show that training with a sequence parallel approach can match standard sequential model training in terms of convergence. Furthermore, we show that our sequence parallel approach has the capability to support training of models that would not be possible on standard computing resources.

Lyngaas, Isaac↗
NASA NTRS1990

Deformation and fracture of aluminum-lithium alloys: The effect of dissolved hydrogen

The effects of dissolved hydrogen on the mechanical properties of 2090 and 2219 alloys are studied. The work done during this semi-annual period consists of the hydrogen charging study and some preliminary mechanical tests. Prior to SIMS analysis, several potentiostatic and galvanostatic experiments were performed for various times (going from 10 minutes to several hours) in the cathodic zone, and for the two aqueous solutions: 0.04N of HCl and 0.1N NaOH both combined with a small amount of As2O3. A study of the surface damage was conducted in parallel with the charging experiments. Those tests were performed to choose the best charging conditions without surface damage. Disk rupture tests and tensile tests are part of the study designed to investigate the effect of temperature, surface roughness, strain rate, and environment on the fracture behavior. The importance of the roughness and environment were shown using the disk rupture test as well as the importance of the strain rate under hydrogen environment. The tensile tests, without hydrogen effects, have not shown significant differences between low and room temperature.

Rivet, F. C.↗
NASA NTRS1996

The " Swarm of Ants vs. Herd of Elephants" Debated Revisited: Performance Measurements of PVM-Overflow Across a Wide Spectrum of Architectures

The Gorden Bell Prizes given out at Supercomputing every year includes at least two catergories: performance (highest GFLOP count) and price-performance (GFLOP/million $$) for real applications. In the past five years, the winners of the price-performance categories all came from networks of work-stations. This reflects three important facts: 1. supercomputers are still too expensive for the masses; 2. achieving high performance for real applications takes real work; and, most importantly; 3. it is possible to obtain acceptable performance for certain real applications on network of work stations. With the continued advance of network technology as well as increased performance of "desktop" workstation, the "Swarm of Ants vs. Herd of Elephants" debate, which began with vector multiprocessors (VPPs) against SIMD type multiprocessors (e.g. CM2), is now recast as VPPs against Symetric Multiprocessors (SMPs, e.g. SGI PowerChallenge). This paper reports on performance studies we performed solving a large scale (2-million grid pt.s) CFD problem involving a Boeing 747 based on a parallel version of OVERFLOW that utilizes message passing on PVM. A performance monitoring tool developed under NASA HPCC, called AIMS, was used to instrument and analyze the the performance data thus obtained. We plan to compare its performance data obtained across a wide spectrum of architectures including: the Cray C90, IBM/SP2, SGI/Power Challenge Cluster, to a group of workstations connected over a simple network. The metrics of comparison includes speed-up, price-performance, throughput, and turn-around time. We also plan to present a plan of attack for various issues that will make the execution of Grand Challenge Applications across the Global Information Infrastructure a reality.

Yan, Jerry C.↗
NASA NTRS2012

Plume-Free Stream Interaction Heating Effects During Orion Crew Module Reentry

During reentry of the Orion Crew Module (CM), vehicle attitude control will be performed by firing reaction control system (RCS) thrusters. Simulation of RCS plumes and their interaction with the oncoming flow has been difficult for the analysis community due to the large scarf angles of the RCS thrusters and the unsteady nature of the Orion capsule backshell environments. The model for the aerothermal database has thus relied on wind tunnel test data to capture the heating effects of thruster plume interactions with the freestream. These data are only valid for the continuum flow regime of the reentry trajectory. A Direct Simulation Monte Carlo (DSMC) analysis was performed to study the vehicle heating effects that result from the RCS thruster plume interaction with the oncoming freestream flow at high altitudes during Orion CM reentry. The study was performed with the DSMC Analysis Code (DAC). The inflow boundary conditions for the jets were obtained from Data Parallel Line Relaxation (DPLR) computational fluid dynamics (CFD) solutions. Simulations were performed for the roll, yaw, pitch-up and pitch-down jets at altitudes of 105 km, 125 km and 160 km as well as vacuum conditions. For comparison purposes (see Figure 1), the freestream conditions were based on previous DAC simulations performed without active RCS to populate the aerodynamic database for the Orion CM. Other inputs to the analysis included a constant Orbital reentry velocity of 7.5 km/s and angle of attack of 160 degrees. The results of the study showed that the interaction effects decrease quickly with increasing altitude. Also, jets with highly scarfed nozzles cause more severe heating compared to the nozzles with lower scarf angles. The difficulty of performing these simulations was based on the maximum number density and the ratio of number densities between the freestream and the plume for each simulation. The lowest altitude solutions required a substantial amount of computational resources (up to 1800 processors) to simulate approximately 2 billion molecules for the refined (adapted) solutions.

Marichalar, J.↗
DOE OSTI2021

Tusas: A fully implicit parallel approach for coupled phase-field equations

In this study, we develop a fully-coupled, fully-implicit approach for phase-field modeling of solidification in metals and alloys. Predictive simulation of solidification in pure metals and metal alloys remains a significant challenge in the field of materials science, as microstructure formation during the solidification process plays a critical role in the properties and performance of the solid material. Our simulation approach consists of a finite element spatial discretization of the fully-coupled nonlinear system of partial differential equations at the microscale, which is treated implicitly in time with a preconditioned Jacobian-free Newton-Krylov method. The approach is algorithmically scalable as well as efficient due to an effective preconditioning strategy based on algebraic multigrid and block factorization. We implement this approach in the open-source Tusas framework, which is a general, flexible tool developed in C++ for solving coupled systems of nonlinear partial differential equations. The performance of our approach is analyzed in terms of algorithmic scalability and efficiency, while the computational performance of Tusas is presented in terms of parallel scalability and efficiency on emerging heterogeneous architectures. We demonstrate that modern algorithms, discretizations, and computational science, and heterogeneous hardware provide a robust route for predictive phase-field simulation of microstructure evolution during additive manufacturing.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1992

Automatic data partitioning on distributed memory multicomputers

Distributed-memory parallel computers are increasingly being used to provide high levels of performance for scientific applications. Unfortunately, such machines are not very easy to program. A number of research efforts seek to alleviate this problem by developing compilers that take over the task of generating communication. The communication overheads and the extent of parallelism exploited in the resulting target program are determined largely by the manner in which data is partitioned across different processors of the machine. Most of the compilers provide no assistance to the programmer in the crucial task of determining a good data partitioning scheme. A novel approach is presented, the constraints-based approach, to the problem of automatic data partitioning for numeric programs. In this approach, the compiler identifies some desirable requirements on the distribution of various arrays being referenced in each statement, based on performance considerations. These desirable requirements are referred to as constraints. For each constraint, the compiler determines a quality measure that captures its importance with respect to the performance of the program. The quality measure is obtained through static performance estimation, without actually generating the target data-parallel program with explicit communication. Each data distribution decision is taken by combining all the relevant constraints. The compiler attempts to resolve any conflicts between constraints such that the overall execution time of the parallel program is minimized. This approach has been implemented as part of a compiler called Paradigm, that accepts Fortran 77 programs, and specifies the partitioning scheme to be used for each array in the program. We have obtained results on some programs taken from the Linpack and Eispack libraries, and the Perfect Benchmarks. These results are quite promising, and demonstrate the feasibility of automatic data partitioning for a significant class of scientific application programs with regular computations.

Gupta, Manish↗
DOE OSTI2022

Performance Enhancement of APW+lo Calculations by Simplest Separation of Concerns

Full-potential linearized augmented plane wave (LAPW) and APW plus local orbital (APW+lo) codes differ widely in both their user interfaces and in capabilities for calculations and analysis beyond their common central task of all-electron solution of the Kohn–Sham equations. However, that common central task opens a possible route to performance enhancement, namely to offload the basic LAPW/APW+lo algorithms to a library optimized purely for that purpose. To explore that opportunity, we have interfaced the Exciting-Plus (“EP”) LAPW/APW+lo DFT code with the highly optimized SIRIUS multi-functional DFT package. This simplest realization of the separation of concerns approach yields substantial performance over the base EP code via additional task parallelism without significant change in the EP source code or user interface. We provide benchmarks of the interfaced code against the original EP using small bulk systems, and demonstrate performance on a spin-crossover molecule and magnetic molecule that are of size and complexity at the margins of the capability of the EP code itself.

Zhang, Long↗
NASA NTRS1993

Infrared sensors and systems for enhanced vision/autonomous landing applications

There exists a large body of data spanning more than two decades, regarding the ability of infrared imagers to 'see' through fog, i.e., in Category III weather conditions. Much of this data is anecdotal, highly specialized, and/or proprietary. In order to determine the efficacy and cost effectiveness of these sensors under a variety of climatic/weather conditions, there is a need for systematic data spanning a significant range of slant-path scenarios. These data should include simultaneous video recordings at visible, midwave (3-5 microns), and longwave (8-12 microns) wavelengths, with airborne weather pods that include the capability of determining the fog droplet size distributions. Existing data tend to show that infrared is more effective than would be expected from analysis and modeling. It is particularly more effective for inland (radiation) fog as compared to coastal (advection) fog, although both of these archetypes are oversimplifications. In addition, as would be expected from droplet size vs wavelength considerations, longwave outperforms midwave, in many cases by very substantial margins. Longwave also benefits from the higher level of available thermal energy at ambient temperatures. The principal attraction of midwave sensors is that staring focal plane technology is available at attractive cost-performance levels. However, longwave technology such as that developed at FLIR Systems, Inc. (FSI), has achieved high performance in small, economical, reliable imagers utilizing serial-parallel scanning techniques. In addition, FSI has developed dual-waveband systems particularly suited for enhanced vision flight testing. These systems include a substantial, embedded processing capability which can perform video-rate image enhancement and multisensor fusion. This is achieved with proprietary algorithms and includes such operations as real-time histograms, convolutions, and fast Fourier transforms.

Kerr, J. Richard↗
NASA NTRS1990

Assignment Of Finite Elements To Parallel Processors

Elements assigned approximately optimally to subdomains. Mapping algorithm based on simulated-annealing concept used to minimize approximate time required to perform finite-element computation on hypercube computer or other network of parallel data processors. Mapping algorithm needed when shape of domain complicated or otherwise not obvious what allocation of elements to subdomains minimizes cost of computation.

Salama, Moktar A.↗
NASA NTRS1998

Charon Message-Passing Toolkit for Scientific Computations

The Charon toolkit for piecemeal development of high-efficiency parallel programs for scientific computing is described. The portable toolkit, callable from C and Fortran, provides flexible domain decompositions and high-level distributed constructs for easy translation of serial legacy code or design to distributed environments. Gradual tuning can subsequently be applied to obtain high performance, possibly by using explicit message passing. Charon also features general structured communications that support stencil-based computations with complex recurrences. Through the separation of partitioning and distribution, the toolkit can also be used for blocking of uni-processor code, and for debugging of parallel algorithms on serial machines. An elaborate review of recent parallelization aids is presented to highlight the need for a toolkit like Charon. Some performance results of parallelizing the NAS Parallel Benchmark SP program using Charon are given, showing good scalability.

VanderWijngaart, Rob F.↗
NASA NTRS1976

Solid-propellant rocket motor internal ballistic performance variation analysis, phase 2

The Monte Carlo method was used to investigate thrust imbalance and its first time derivative throughtout the burning time of pairs of solid rocket motors firing in parallel. Results obtained compare favorably with Titan 3 C flight performance data. Statistical correlations of the thrust imbalance at various times with corresponding nominal trace slopes suggest several alternative methods of predicting thrust imbalance. The effect of circular-perforated grain deformation on internal ballistics is discussed, and a modified design analysis computer program which permits such an evaluation is presented. Comparisons with SRM firings indicate that grain deformation may account for a portion of the so-called scale factor on burning rate between large motors and strand burners or small ballistic test motors. Thermoelastic effects on burning rate are also investigated. Burning surface temperature is calculated by coupling the solid phase energy equation containing a strain rate term with a model of gas phase combustion zone using the Zeldovich-Novozhilov technique. Comparisons of solutions with and without the strain rate term indicate a small but possibly significant effect of the thermoelastic coupling.

Sforzini, R. H.↗
NASA NTRS2012

Production Level CFD Code Acceleration for Hybrid Many-Core Architectures

In this work, a novel graphics processing unit (GPU) distributed sharing model for hybrid many-core architectures is introduced and employed in the acceleration of a production-level computational fluid dynamics (CFD) code. The latest generation graphics hardware allows multiple processor cores to simultaneously share a single GPU through concurrent kernel execution. This feature has allowed the NASA FUN3D code to be accelerated in parallel with up to four processor cores sharing a single GPU. For codes to scale and fully use resources on these and the next generation machines, codes will need to employ some type of GPU sharing model, as presented in this work. Findings include the effects of GPU sharing on overall performance. A discussion of the inherent challenges that parallel unstructured CFD codes face in accelerator-based computing environments is included, with considerations for future generation architectures. This work was completed by the author in August 2010, and reflects the analysis and results of the time.

Duffy, Austen C.↗
DOE OSTI2020

EQSIM—A multidisciplinary framework for fault-to-structure earthquake simulations on exascale computers part I: Computational models and workflow

Computational simulations have become central to the seismic analysis and design of major infrastructure over the past several decades. Most major structures are now “proof tested” virtually through representative simulations of earthquake-induced response. More recently, with the advancement of high-performance computing (HPC) platforms and the associated massively parallel computational ecosystems, simulation is beginning to play a role in increased understanding and prediction of ground motions for earthquake hazard assessments. However, the computational requirements for regional-scale geophysics-based ground motion simulations are extreme, which has restricted the frequency resolution of direct simulations and limited the ability to perform the large number of simulations required to numerically explore the problem parametric space. In this article, recent developments toward an integrated, multidisciplinary earth science-engineering computational framework for the regional-scale simulation of both ground motions and resulting structural response are described with a particular emphasis on advancing simulations to frequencies relevant to engineered systems. This multidisciplinary computational development is being carried out as part of the US Department of Energy (DOE) Exascale Computing Project with the goal of achieving a computational framework poised to exploit emerging DOE exaflop computer platforms scheduled for the 2022–2023 timeframe.

58 GEOSCIENCES↗
DOE OSTI2024

Experimental Characterization Test of a Grid-Forming Inverter for Microgrid Applications

Standardized experimental testing protocols for grid forming (GFM) inverters to ensure expected operation under both normal and contingency conditions do not exist. Such protocols increase the confidence of system owner/operators that an inverter deployed in a proposed system will engage in typical behaviors to ensure interoperability with other units and ancillary equipment (e.g. protection equipment). This paper presents systematic and comprehensive test protocols to evaluate the performance of GFM inverters under the following operational configurations: islanded operation, heterogeneous islanded operation (parallel with a synchronous generator), grid-connected operation, and transition operation. A commercial GFM inverter is used to verify the test protocols and to understand the inverter's performance and functionalities. In particular, required configuration and tuning of the inverter will be explained in the full paper to enrich the testing protocol.

black start↗
NASA NTRS1991

S-R compatibility effects with orthogonal stimulus and response dimensions

An experiment was conducted to assess the relative influence of several factors on performance with orthogonal stimulus-response arrays. Subjects responded to the onset of one of three aligned light circles with a press of one of three aligned response keys. The response array was aligned parallel, angled, or orthogonal to the stimulus array. The results indicated that performance with orthogonal arrays is worse than with parallel or angled S-R arrays. For the orthogonal arrangements, the results also indicate that each hand prefers a mapping directly opposite to the other hand, and that this mapping reverses when the orientation of the stimulus and response arrays are transposed. In addition, the results also revealed that the relative costs of orthogonal S-R arrangements are somewhat attenuated when the assigned mapping associates (i.e., colocates) a given display with its closest control.

Andre, Anthony D.↗