Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel processors”

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 469 records · Page 26

Hybrid and Parallel Domain-Decomposition Methods Development to Enable Monte Carlo for Reactor Analyses

This paper describes code and methods development at the Oak Ridge National Laboratory focused on enabling high-fidelity, large-scale reactor analyses with Monte Carlo (MC). Current state-of-the-art tools and methods used to perform real commercial reactor analyses have several undesirable features, the most significant of which is the non-rigorous spatial decomposition scheme. Monte Carlo methods, which allow detailed and accurate modeling of the full geometry and are considered the gold standard for radiation transport solutions, are playing an ever-increasing role in correcting and/or verifying the deterministic, multi-level spatial decomposition methodology in current practice. However, the prohibitive computational requirements associated with obtaining fully converged, system-wide solutions restrict the role of MC to benchmarking deterministic results at a limited number of state-points for a limited number of relevant quantities. The goal of this research is to change this paradigm by enabling direct use of MC for full-core reactor analyses. The most significant of the many technical challenges that must be overcome are the slow, non-uniform convergence of system-wide MC estimates and the memory requirements associated with detailed solutions throughout a reactor (problems involving hundreds of millions of different material and tally regions due to fuel irradiation, temperature distributions, and the needs associated with multi-physics code coupling). To address these challenges, our research has focused on the development and implementation of (1) a novel hybrid deterministic/MC method for determining high-precision fluxes throughout the problem space in k-eigenvalue problems and (2) an efficient MC domain-decomposition (DD) algorithm that partitions the problem phase space onto multiple processors for massively parallel systems, with statistical uncertainty estimation. The hybrid method development is based on an extension of the FW-CADIS method, which attempts to achieve uniform statistical uncertainty throughout a designated problem space. The MC DD development is being implemented in conjunction with the Denovo deterministic radiation transport package to have direct access to the 3-D, massively parallel discrete-ordinates solver (to support the hybrid method) and the associated parallel routines and structure. This paper describes the hybrid method, its implementation, and initial testing results for a realistic 2-D quarter core pressurized-water reactor model and also describes the MC DD algorithm and its implementation.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Hybrid and Parallel Domain-Decomposition Methods Development to Enable Monte Carlo for Reactor Analyses

This paper describes code and methods development at the Oak Ridge National Laboratory focused on enabling high-fidelity, large-scale reactor analyses with Monte Carlo (MC). Current state-of-the-art tools and methods used to perform ''real'' commercial reactor analyses have several undesirable features, the most significant of which is the non-rigorous spatial decomposition scheme. Monte Carlo methods, which allow detailed and accurate modeling of the full geometry and are considered the ''gold standard'' for radiation transport solutions, are playing an ever-increasing role in correcting and/or verifying the deterministic, multi-level spatial decomposition methodology in current practice. However, the prohibitive computational requirements associated with obtaining fully converged, system-wide solutions restrict the role of MC to benchmarking deterministic results at a limited number of state-points for a limited number of relevant quantities. The goal of this research is to change this paradigm by enabling direct use of MC for full-core reactor analyses. The most significant of the many technical challenges that must be overcome are the slow, non-uniform convergence of system-wide MC estimates and the memory requirements associated with detailed solutions throughout a reactor (problems involving hundreds of millions of different material and tally regions due to fuel irradiation, temperature distributions, and the needs associated with multi-physics code coupling). To address these challenges, our research has focused on the development and implementation of (1) a novel hybrid deterministic/MC method for determining high-precision fluxes throughout the problem space in k-eigenvalue problems and (2) an efficient MC domain-decomposition (DD) algorithm that partitions the problem phase space onto multiple processors for massively parallel systems, with statistical uncertainty estimation. The hybrid method development is based on an extension of the FW-CADIS method, which attempts to achieve uniform statistical uncertainty throughout a designated problem space. The MC DD development is being implemented in conjunction with the Denovo deterministic radiation transport package to have direct access to the 3-D, massively parallel discrete-ordinates solver (to support the hybrid method) and the associated parallel routines and structure. This paper describes the hybrid method, its implementation, and initial testing results for a realistic 2-D quarter core pressurized-water reactor model and also describes the MC DD algorithm and its implementation.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Parallel Agent-Based Simulations on Clusters of GPUs and Multi-Core Processors

An effective latency-hiding mechanism is presented in the parallelization of agent-based model simulations (ABMS) with millions of agents. The mechanism is designed to accommodate the hierarchical organization as well as heterogeneity of current state-of-the-art parallel computing platforms. We use it to explore the computation vs. communication trade-off continuum available with the deep computational and memory hierarchies of extant platforms and present a novel analytical model of the tradeoff. We describe our implementation and report preliminary performance results on two distinct parallel platforms suitable for ABMS: CUDA threads on multiple, networked graphical processing units (GPUs), and pthreads on multi-core processors. Message Passing Interface (MPI) is used for inter-GPU as well as inter-socket communication on a cluster of multiple GPUs and multi-core processors. Results indicate the benefits of our latency-hiding scheme, delivering as much as over 100-fold improvement in runtime for certain benchmark ABMS application scenarios with several million agents. This speed improvement is obtained on our system that is already two to three orders of magnitude faster on one GPU than an equivalent CPU-based execution in a popular simulator in Java. Thus, the overall execution of our current work is over four orders of magnitude faster when executed on multiple GPUs.

97 MATHEMATICS AND COMPUTING↗

Linearly Scaling 3D Fragment Method for Large-Scale Electronic Structure Calculations

We present a new linearly scaling three-dimensional fragment (LS3DF) method for large scale ab initio electronic structure calculations. LS3DF is based on a divide-and-conquer approach, which incorporates a novel patching scheme that effectively cancels out the artificial boundary effects due to the subdivision of the system. As a consequence, the LS3DF program yields essentially the same results as direct density functional theory (DFT) calculations. The fragments of the LS3DF algorithm can be calculated separately with different groups of processors. This leads to almost perfect parallelization on tens of thousands of processors. After code optimization, we were able to achieve 35.1 Tflop/s, which is 39percent of the theoretical speed on 17,280 Cray XT4 processor cores. Our 13,824-atom ZnTeO alloy calculation runs 400 times faster than a direct DFTcalculation, even presuming that the direct DFT calculation can scale well up to 17,280 processor cores. These results demonstrate the applicability of the LS3DF method to material simulations, the advantage of using linearly scaling algorithms over conventional O(N3) methods, and the potential for petascale computation using the LS3DF method.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

A Parallel Ghosting Algorithm for The Flexible Distributed Mesh Database

Critical to the scalability of parallel adaptive simulations are parallel control functions including load balancing, reduced inter-process communication and optimal data decomposition. In distributed meshes, many mesh-based applications frequently access neighborhood information for computational purposes which must be transmitted efficiently to avoid parallel performance degradation when the neighbors are on different processors. This article presents a parallel algorithm of creating and deleting data copies, referred to as ghost copies, which localize neighborhood data for computation purposes while minimizing inter-process communication. The key characteristics of the algorithm are: (1) It can create ghost copies of any permissible topological order in a 1D, 2D or 3D mesh based on selected adjacencies. (2) It exploits neighborhood communication patterns during the ghost creation process thus eliminating all-to-all communication. (3) For applications that need neighbors of neighbors, the algorithm can create n number of ghost layers up to a point where the whole partitioned mesh can be ghosted. Strong and weak scaling results are presented for the IBM BG/P and Cray XE6 architectures up to a core count of 32,768 processors. The algorithm also leads to scalable results when used in a parallel super-convergent patch recovery error estimator, an application that frequently accesses neighborhood data to carry out computation.

Mubarak, Misbah↗

Flexible User-Defined Domain Decomposition in Kilometer-Scale E3SM Land Model Simulation

The Energy Exascale Earth System Model (E3SM) Land Model (ELM) has been extended to kilometer-scale (km-ELM) resolutions, enabling high-fidelity simulations of terrestrial processes at 1 km x 1 km grid spacing. In ELM, domain decomposition partitions the computational domain across processors, ensuring efficient parallel execution. Currently, round-robin decomposition is applied, providing a straightforward way to distribute computational workload. As ELM continues evolving at the kilometer-scale (km-scale), particularly with integrating lateral flow modeling, decomposition strategies must also account for the increased workload and data movement. This paper introduces a flexible user-defined domain decomposition framework, allowing users to customize domain partitioning based on application requirements. The impact of different decomposition strategies is evaluated across various applications concerning computation, communication, and I/O. Results demonstrate that while 1D partitioning yields superior I/O performance, k-nearest neighbors (KNN) clustering effectively reduces inter-process communication overhead. This study lays the groundwork for scalable partitioning in large-scale land surface simulations, enhancing next-generation Earth system modeling.

Wang, Dali [ORNL] (ORCID:0000000168065108)↗

LES of an ignition sequence in a gas turbine engine

Being able to ignite or reignite a gas turbine engine in a cold and rarefied atmosphere is a critical issue for many manufacturers. From a fundamental point of view, the ignition of the first burner and the flame propagation from one burner to another are phenomena that are usually not studied. The present work is a large eddy simulation (LES) of these phenomena. To simulate a complete ignition sequence in an annular chamber, LES has been applied to the full 360 geometry, including 18 burners. This geometry corresponds to a real gas turbine chamber. Massively parallel computing (700 processors on a Cray XT3 machine) was essential to perform such a large calculation. Results show that liquid fuel injection has a strong influence on the ignition times. Moreover, the rate of flame progress from burner to burner is much higher than the turbulent flame speed due to a major effect of thermal expansion. This flame speed is also strongly modified by the main burner aerodynamics due to the swirled injection. Finally, the variability of the combustor sectors and quadrant ignition times is highlighted. (author)

42 ENGINEERING↗

Ab initio wavenumber accurate spectroscopy : {sup 1}CH{sub 2} and HCN vibrational levels on automatically generated IMLS potential energy surfaces.

We report here calculated J = 0 vibrational frequencies for {sup 1}CH{sub 2} and HCN with root-mean-square error relative to available measurements of 2.0 cm{sup -1} and 3.2 cm{sup -1}, respectively. These results are obtained with DVR calculations with a dense grid on ab initio potential energy surfaces (PESs). The ab initio electronic structure calculations employed are Davidson-corrected MRCI calculations with double-, triple-, and quadruple-{zeta} basis sets extrapolated to the complete basis set (CBS) limit. In the {sup 1}CH{sub 2} case, Full CI tests of the Davidson correction at small basis set levels lead to a scaling of the correction with the bend angle that can be profitably applied at the CBS limit. Core-valence corrections are added derived from CCSD(T) calculations with and without frozen cores. Relativistic and non-Born-Oppenheimer corrections are available for HCN and were applied. CBS limit CCSD(T) and CASPT2 calculations with the same basis sets were also tried for HCN. The CCSD(T) results are noticeably less accurate than the MRCI results while the CASPT2 results are much poorer. The PESs were generated automatically using the local interpolative moving least-squares method (L-IMLS). A general triatomic code is described where the L-IMLS method is interfaced with several common electronic structure packages. All PESs were computed with this code running in parallel on eight processors. The L-IMLS method provides global and local fitting error measures important in automatically growing the PES from initial ab initio seed points. The reliability of this approach was tested for {sup 1}CH{sub 2} by comparing DVR-calculated vibrational levels on an L-IMLS ab initio surface with levels generated by an explicit ab initio calculation at each DVR grid point. For all levels ({approx}200) below 20000 cm{sup -1}, the mean unsigned difference between the levels of these two calculations was 0.1 cm{sup -1}, consistent with the L-IMLS estimated mean unsigned fitting error of 0.3 cm{sup -1}. All L-IMLS PESs used in this work have comparable mean unsigned fitting errors, implying that fitting errors have a negligible role in the final errors of the computed vibrational levels with experiment. Less than 500 ab initio calculations of the energy and gradients are required to achieve this level of accuracy.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Analyzing the Impact of System Reliability Events on Applications in the Titan Supercomputer

Extreme-scale computing systems employ Reliability, Availability and Serviceability (RAS) mechanisms and infrastructure to log events from multiple system components. In this paper, we analyze RAS logs in conjunction with the application placement and scheduling database, in order to understand the impact of common RAS events on application performance. This study conducted on the records of about 2 million applications executed on Titan supercomputer provides important insights for system users, operators and computer science researchers. Specifically, we investigate the impact of RAS events on application performance and its variability by comparing cases where events are recorded with corresponding cases where no events are recorded. Such a statistical investigation is possible since we observed that system users tend to execute their applications multiple times. Our analysis reveals that most RAS events do impact application performance, although not always. We also find that different system components affect application performance differently. In particular, our investigation includes the following components: parallel file system processor, memory, graphics processing units, system and user software issues. Our work establishes the importance of providing feedback to system users for increasing operational efficiency of extreme-scale systems.

Ashraf, Rizwan↗

AnisWave2D

AnisWave2D is a 2D finite-difference code for simulating seismic wave propagation in fully anisotropic materials. The code is implemented to run in parallel over multiple processors and is fully portable. A mesh refinement algorithm has been utilized to allow the grid-spacing to be tailored to the velocity model, avoiding the over-sampling of high-velocity materials that usually occurs in fixed-grid schemes.

Toomey, Aoife↗

EJFAT: Towards Intelligent Compute Destination Load Balancing

To handle increased data flow, Jefferson Lab (JLab) is partnering with ESnet for development of an AI/ML directed compute work Load Balancer (LB) of UDP streamed data. The LB is FPGA based featuring dynamically configurable, low latency and high throughput destination address switching. The LB provides integration of edge and core computing to support JLab experimental programs, the Electron-Ion Collider, as well as data centers of the future. In the ESnet/JLab FPGA Accelerated Transport (EJFAT) initiative, the function of the LB Data Plane (DP) is to redirect data streams to selectable (but unknown to sender) destination hosts based on current worload and within that host to destination ports as a function of sub- stream id. This effects hierarchical scaling, first across compute machines for processing over a series of events and second, across ports so different data source sub-streams may be assigned to different processors for further parallelization. The LB Control Plane (CP) programs the DP using compute farm telemetry to direct and balance workloads across a compute cluster as the operating conditions require. While Proportional/Integrative/Derivative (PID) controllers are often seen in similar applications, here we investigate the feasibility of a Reinforcement Learning (RL) based schedule manager running in the CP to provide dynamic updates to the DP scheduling policy.

Lawrence, David↗

Burnup calculations for KIPT accelerator driven subcritical facility using Monte Carlo computer codes-MCB and MCNPX.

Argonne National Laboratory (ANL) of USA and Kharkov Institute of Physics and Technology (KIPT) of Ukraine have been collaborating on the conceptual design development of an electron accelerator driven subcritical (ADS) facility, using the KIPT electron accelerator. The neutron source of the subcritical assembly is generated from the interaction of 100 KW electron beam with a natural uranium target. The electron beam has a uniform spatial distribution and electron energy in the range of 100 to 200 MeV. The main functions of the subcritical assembly are the production of medical isotopes and the support of the Ukraine nuclear power industry. Neutron physics experiments and material structure analyses are planned using this facility. With the 100 KW electron beam power, the total thermal power of the facility is {approx}375 kW including the fission power of {approx}260 kW. The burnup of the fissile materials and the buildup of fission products reduce continuously the reactivity during the operation, which reduces the neutron flux level and consequently the facility performance. To preserve the neutron flux level during the operation, fuel assemblies should be added after long operating periods to compensate for the lost reactivity. This process requires accurate prediction of the fuel burnup, the decay behavior of the fission produces, and the introduced reactivity from adding fresh fuel assemblies. The recent developments of the Monte Carlo computer codes, the high speed capability of the computer processors, and the parallel computation techniques made it possible to perform three-dimensional detailed burnup simulations. A full detailed three-dimensional geometrical model is used for the burnup simulations with continuous energy nuclear data libraries for the transport calculations and 63-multigroup or one group cross sections libraries for the depletion calculations. Monte Carlo Computer code MCNPX and MCB are utilized for this study. MCNPX transports the electrons and the produced neutrons and photons but the current version of MCNPX doesn't support depletion/burnup calculation of the subcritical system with the generated neutron source from the target. MCB can perform neutron transport and burnup calculation for subcritical system using external neutron source, however it cannot perform electron transport calculations. To solve this problem, a hybrid procedure is developed by coupling these two computer codes. The user tally subroutine of MCNPX is developed and utilized to record the information of the each generated neutron from the photonuclear reactions resulted from the electron beam interactions. MCB reads the recorded information of each generated neutron thorough the user source subroutine. In this way, the neutron source generated by electron reactions could be utilized in MCB calculations, without the need for MCB to transport the electrons. Using the source subroutines, MCB could get the external neutron source, which is prepared by MCNPX, and perform depletion calculation for the driven subcritical facility.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Optical production systems using neural networks and symbolic substitution

Two optical implementations of production systems are advanced. The production systems operate on a knowledge base where facts and rules are encoded as formulas in propositional calculus. The first implementation is a binary neural network. An analog neural network is used to include reasoning with uncertainties. The second implementation uses a new optical symbolic substitution correlator. This implementation is useful when a set of similar situations has to be handled in parallel on one processor.

Botha, Elizabeth↗

Hierarchical Poly Tree computer architectures defined by computational multidisciplinary mechanics

This paper will develop an alternative computer architecture called the Poly Tree. Based on the requirements of computational mechanics and the concept of hierarchical substructuring, the paper will explore the development of problem-dependent parallel networks of processors which will enable significant, often superlinear, speed enhancements; provide a logical/efficient framework for linear/nonlinear and transient structural mechanics problems; and provide a logical framework from which to apply model reduction procedures. In addition, the paper will explore optimal processor arrangements which define the overall system granularity. Consideration will also be given to system I/O requirements.

Padovan, Joe↗

Memory management in traceback Viterbi decoders

The new Viterbi decoder for long constraint length codes, under development for the Deep Space Network, stores path information according to an algorithm called traceback. The details of a particular implementation of this algorithm, based on three memory buffers, are described. The penalties in increased storage requirement and longer decoding delay are offset by the reduced amount of data that needs to be exchanged between processors, in a parallel architecture decoder.

Collins, O.↗

A Strassen-Newton algorithm for high-speed parallelizable matrix inversion

Techniques are described for computing matrix inverses by algorithms that are highly suited to massively parallel computation. The techniques are based on an algorithm suggested by Strassen (1969). Variations of this scheme use matrix Newton iterations and other methods to improve the numerical stability while at the same time preserving a very high level of parallelism. One-processor Cray-2 implementations of these schemes range from one that is up to 55 percent faster than a conventional library routine to one that is slower than a library routine but achieves excellent numerical stability. The problem of computing the solution to a single set of linear equations is discussed, and it is shown that this problem can also be solved efficiently using these techniques.

Bailey, David H.↗

Mutual exclusion

Almost all computers today operate as part of a network, where they assist people in coordinating actions. Sometimes what appears to be a single computer is actually a network of cooperating computers; e.g., some supercomputers consist of many processors operating in parallel and exchanging synchronization signals. One of the most fundamental requirements in all these systems is that certain operations be indivisible: the steps of one must not be interleaved with the steps of another. Two approaches were designed to implement this requirement, one based on central locks and the other on distributed order tickets. Practicing scientists and engineers need to come to be familiar with these methods.

Denning, Peter J.↗

Computational requirements for on-orbit identification of space systems

For the future space systems, on-orbit identification (ID) capability will be required to complement on-orbit control, due to the fact that the dynamics of large space structures, spacecrafts, and antennas will not be known sufficiently from ground modeling and testing. The computational requirements for ID of flexible structures such as the space station (SS) or the large deployable reflectors (LDR) are however, extensive due to the large number of modes, sensors, and actuators. For these systems the ID algorithm operations need not be computed in real-time, only in near real-time, or an appropriate mission time. Consequently the space systems will need advanced processors and efficient parallel processing algorithm design and architectures to implement the identification algorithms in near real-time. The MAX computer currently being developed may handle such computational requirements. The purpose is to specify the on-board computational requirements for dynamic and static identification for large space structures. The computational requirements for six ID algorithms are presented in the context of three examples: the JPL/AFAL ground antenna facility, the space station (SS), and the large deployable reflector (LDR).

Hadaegh, Fred Y.↗