Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Parallel algorithms.”

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

Investigation of Portable Event-Based Monte Carlo Transport Using the NVIDIA Thrust Library

Power consumption considerations are driving future high performance computing platforms toward many-core computing architectures. Los Alamos National Laboratory's Trinity machine, available in 2016, will use both Intel Xeon Haswell processors and Intel Xeon Phi Knights Landing many integrated core (MIC) architecture coprocessors. Lawrence Livermore National Laboratory's Sierra machine, available in 2018, will use an IBM PowerPC architecture along with Nvidia graphics processing unit (GPU) architecture accelerators. These different advanced architectures make the computing landscape in upcoming years complex. Traditional approaches to Monte Carlo transport do not work efficiently on these new computing platforms. MIC architectures require vectorization to operate efficiently, and vectorization is difficult to achieve in Monte Carlo transport. GPU architectures require additional code to explicitly use the hardware, requiring significant code changes or hardware specific branches in the source code. A significant challenge for Monte Carlo transport projects is to simultaneously support within a single source code base efficient simulations for both the current generation of architectures and the different advanced computing architectures. In order to address these challenges, two important changes are typically required: a new algorithmic approach for solving Monte Carlo transport, and explicit use of hardware specific software. In this paper, we describe initial research investigations of an event-based Monte Carlo transport algorithm implemented using the Nvidia Thrust library on a GPU for a Monte Carlo test code. The event-based algorithm targets many-core architectures by increasing SIMD (single instruction multiple data) parallelism, while Thrust potentially provides portable performance by allowing one source code base to compile code targeted for both CPUs and GPUs. We described preliminary investigations of portable event-based Monte Carlo algorithms implemented using the Nvidia Thrust library in a research Monte Carlo test code. We found that an explicit CUDA implementation of an event-based Monte Carlo algorithm performed significantly more efficiently than a Thrust implementation on GPU platforms, most likely as a result of additional flexibility in access to different memory spaces on the GPU. Additionally, we showed that on GPU platforms and at large enough problem sizes the event-based implementations perform more efficiently than the serial history-based implementation running on the host CPU. While investigating this problem, we also discovered that the performance of the event-based algorithm is affected by what tallies are being used. A zonal scalar flux tally requires atomic operations that significantly impacted the performance of the code, in some cases producing slowdowns instead of speedups. We decided to remove the tally in order to focus on the effectiveness of the event-based algorithm. Future work will be required to research more effective ways of handling such tallies. Additionally, we would like to consider new ways for optimizing both the Thrust and CUDA versions, in order to see how much performance we have yet to achieve. The potential trade-off between portability and performance was demonstrated in this investigation. Thrust provides both CPU and GPU versions of the code in one code base, but it does so at a cost. We discovered approximately a factor of 2-5 performance difference between Thrust and CUDA on each GPU platform we tested. Thrust provides a tool to access the GPUs with less effort and specialization, but it does so by giving up fine grained control where extra performance can be found for this application. Future work will be required to determine whether this performance differential between Thrust and CUDA can be reduced for event-based Monte Carlo transport. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Parallel time integration software

This package implements an optimal-scaling multigrid solver for the (non) linear systems that arise from the discretization of problems with evolutionary behavior. Typically, solution algorithms for evolution equations are based on a time-marching approach, solving sequentially for one time step after the other. Parallelism in these traditional time-integrarion techniques is limited to spatial parallelism. However, current trends in computer architectures are leading twards system with more, but not faster. processors. Therefore, faster compute speeds must come from greater parallelism. One approach to achieve parallelism in time is with multigrid, but extending classical multigrid methods for elliptic poerators to this setting is a significant achievement. In this software, we implement a non-intrusive, optimal-scaling time-parallel method based on multigrid reduction techniques. The examples in the package demonstrate optimality of our multigrid-reduction-in-time algorithm (MGRIT) for solving a variety of parabolic equations in two and three sparial dimensions. These examples can also be used to show that MGRIT can achieve significant speedup in comparison to sequential time marching on modern architectures.

Source record↗

Efficient Graph Based Assembly of Short-Read Sequences on Hybrid Core Architecture

Advanced architectures can deliver dramatically increased throughput for genomics and proteomics applications, reducing time-to-completion in some cases from days to minutes. One such architecture, hybrid-core computing, marries a traditional x86 environment with a reconfigurable coprocessor, based on field programmable gate array (FPGA) technology. In addition to higher throughput, increased performance can fundamentally improve research quality by allowing more accurate, previously impractical approaches. We will discuss the approach used by Convey?s de Bruijn graph constructor for short-read, de-novo assembly. Bioinformatics applications that have random access patterns to large memory spaces, such as graph-based algorithms, experience memory performance limitations on cache-based x86 servers. Convey?s highly parallel memory subsystem allows application-specific logic to simultaneously access 8192 individual words in memory, significantly increasing effective memory bandwidth over cache-based memory systems. Many algorithms, such as Velvet and other de Bruijn graph based, short-read, de-novo assemblers, can greatly benefit from this type of memory architecture. Furthermore, small data type operations (four nucleotides can be represented in two bits) make more efficient use of logic gates than the data types dictated by conventional programming models.JGI is comparing the performance of Convey?s graph constructor and Velvet on both synthetic and real data. We will present preliminary results on memory usage and run time metrics for various data sets with different sizes, from small microbial and fungal genomes to very large cow rumen metagenome. For genomes with references we will also present assembly quality comparisons between the two assemblers.

59 BASIC BIOLOGICAL SCIENCES↗

Parallel time integration software

This package implements an optimal-scaling multigrid solver for the (non) linear systems that arise from the discretization of problems with evolutionary behavior. Typically, solution algorithms for evolution equations are based on a time-marching approach, solving sequentially for one time step after the other. Parallelism in these traditional time-integrarion techniques is limited to spatial parallelism. However, current trends in computer architectures are leading twards system with more, but not faster. processors. Therefore, faster compute speeds must come from greater parallelism. One approach to achieve parallelism in time is with multigrid, but extending classical multigrid methods for elliptic poerators to this setting is a significant achievement. In this software, we implement a non-intrusive, optimal-scaling time-parallel method based on multigrid reduction techniques. The examples in the package demonstrate optimality of our multigrid-reduction-in-time algorithm (MGRIT) for solving a variety of parabolic equations in two and three sparial dimensions. These examples can also be used to show that MGRIT can achieve significant speedup in comparison to sequential time marching on modern architectures.

Source record↗

DAG-Based Software Frameworks for PDEs

The task-based approach to software and parallelism is well-known and has been proposed as a potential candidate, named the silver model, for exascale software. This approach is not yet widely used in the large-scale multi-core parallel computing of complex systems of partial differential equations. After surveying task-based approaches we investigate how well the Uintah software and an extension named Wasatch fit in the task-based paradigm and how well they perform on large scale parallel computers. The conclusion is that these approaches show great promise for petascale but that considerable algorithmic challenges remain.

Directed Acyclic Graph↗

A NUMERICAL ALGORITHM FOR MODELING MULTIGROUP NEUTRINO-RADIATION HYDRODYNAMICS IN TWO SPATIAL DIMENSIONS

It is now generally agreed that multidimensional, multigroup, neutrino-radiation hydrodynamics (RHD) is an indispensable element of any realistic model of stellar-core collapse, core-collapse supernovae, and proto-neutron star instabilities. We have developed a new, two-dimensional, multigroup algorithm that can model neutrino-RHD flows in core-collapse supernovae. Our algorithm uses an approach similar to the ZEUS family of algorithms, originally developed by Stone and Norman. However, this completely new implementation extends that previous work in three significant ways: first, we incorporate multispecies, multigroup RHD in a flux-limited-diffusion approximation. Our approach is capable of modeling pair-coupled neutrino-RHD, and includes effects of Pauli blocking in the collision integrals. Blocking gives rise to nonlinearities in the discretized radiation-transport equations, which we evolve implicitly in time. We employ parallelized Newton-Krylov methods to obtain a solution of these nonlinear, implicit equations. Our second major extension to the ZEUS algorithm is the inclusion of an electron conservation equation that describes the evolution of electron-number density in the hydrodynamic flow. This permits calculating deleptonization of a stellar core. Our third extension modifies the hydrodynamics algorithm to accommodate realistic, complex equations of state, including those having nonconvex behavior. In this paper, we present a description of our complete algorithm, giving sufficient details to allow others to implement, reproduce, and extend our work. Finite-differencing details are presented in appendices. We also discuss implementation of this algorithm on state-of-the-art, parallel-computing architectures. Finally, we present results of verification tests that demonstrate the numerical accuracy of this algorithm on diverse hydrodynamic, gravitational, radiation-transport, and RHD sample problems. We believe our methods to be of general use in a variety of model settings where radiation transport or RHD is important. Extension of this work to three spatial dimensions is straightforward.

79 ASTRONOMY AND ASTROPHYSICS↗

Graphics processing unit-assisted lossless decompression

Systems and methods for decompressing compressed data that has been compressed by way of a lossless compression algorithm are described herein. In a general embodiment, a graphics processing unit (GPU) is programmed to receive compressed data packets and decompress such packets in parallel. The compressed data packets are compressed representations of an image, and the lossless compression algorithm is a Rice compression algorithm.

97 MATHEMATICS AND COMPUTING↗

Developing highly scalable fluid solvers for enabling multiphysics simulation.

We performed an investigation into explicit algorithms for the simulation of incompressible flows using methods with a finite, but small amount of compressibility added. Such methods include the artificial compressibility method and the lattice-Boltzmann method. The impetus for investigating such techniques stems from the increasing use of parallel computation at all levels (processors, clusters, and graphics processing units). Explicit algorithms have the potential to leverage these resources. In our investigation, a new form of artificial compressibility was derived. This method, referred to as the Entropically Damped Artificial Compressibility (EDAC) method, demonstrated superior results to traditional artificial compressibility methods by damping the numerical acoustic waves associated with these methods. Performance nearing that of the lattice- Boltzmann technique was observed, without the requirement of recasting the problem in terms of particle distribution functions; continuum variables may be used. Several example problems were investigated using a finite-di erence and finite-element discretizations of the EDAC equations. Example problems included lid-driven cavity flow, a convecting Taylor-Green vortex, a doubly periodic shear layer, freely decaying turbulence, and flow over a square cylinder. Additionally, a scalability study was performed using in excess of one million processing cores. Explicit methods were found to have desirable scaling properties; however, some robustness and general applicability issues remained.

Clausen, Jonathan R↗

toolkit computational mesh conceptual model.

The Sierra Toolkit computational mesh is a software library intended to support massively parallel multi-physics computations on dynamically changing unstructured meshes. This domain of intended use is inherently complex due to distributed memory parallelism, parallel scalability, heterogeneity of physics, heterogeneous discretization of an unstructured mesh, and runtime adaptation of the mesh. Management of this inherent complexity begins with a conceptual analysis and modeling of this domain of intended use; i.e., development of a domain model. The Sierra Toolkit computational mesh software library is designed and implemented based upon this domain model. Software developers using, maintaining, or extending the Sierra Toolkit computational mesh library must be familiar with the concepts/domain model presented in this report.

97 MATHEMATICS AND COMPUTING↗

Genetic algorithm optimization of nuclear criticality experiment for reduction of intermediate-energy 239 Pu nuclear data uncertainties

Nuclear criticality experiments are conducted to investigate specific nuclear data important for safe handling and storage of fissile materials, reactor design and operation, and the validation of radiation transport codes. Incorrect or uncertain nuclear data can prohibitively impact operational safety limits, reactor licensing, and predictive simulation capability; therefore, integral measurements from criticality experiments are necessary and should be performed frequently. To maximize the impact of the integral measurements, it is important to consider experiment geometry, material selection, and component dimensions. When taking these considerations into account, the experiment design process becomes iterative and very time intensive. This work utilizes a genetic algorithm to efficiently explore potential nuclear criticality experiment designs for the Laboratory Directed Research & Development project PARADIGM (PARallel Approach of Differential and InteGral Measurements) at Los Alamos National Laboratory. In this paper, the building blocks of the genetic algorithm are discussed in detail, the genetic algorithm methodology is verified, and the genetic algorithm is used to produce three candidate experiment models for the final PARADIGM design. The three candidate models produced by the genetic algorithm consist of copper-reflected assemblies containing 14 repeating units of alumina, graphite, boron, and plutonium plates. Furthermore, in addition to the optimization results, final design considerations are also discussed for designs with a height and/or weight very close to or slightly above assembly machine operational limits.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Development of algorithms for augmenting and replacing conventional process control using reinforcement learning

Here, this work seeks to allow for the online operation and training of model-free reinforcement learning (RL) agents but limit the risk to system equipment and personnel. The parallel implementation of RL alongside more conventional process control (CPC) allows for the RL algorithm to learn from CPC. The past performance of both methods are assessed on a continuous basis allowing for a transition from CPC to RL and, if needed, transitioning back to CPC from RL. This allows for the RL algorithm to slowly and safely assume control of the process without significant degradation in control performance. It is shown that the RL can derive a near optimal policy even when coupled with a suboptimal CPC. It is also demonstrated that the coupled RL-CPC algorithm learns at a faster rate than traditional RL methods of exploration while the algorithm’s performance does not deteriorate below CPC, even when exposed to an unknown operating condition.

30 DIRECT ENERGY CONVERSION↗

Parallel Variable Population Multi-Objective Optimizer (pvpmoo) v1.0

This is a parallel variable population multi-objective optimizer with an adaptive unified differential evolution algorithm or a genetic algorithm. It can also be used for single objective optimization. Some features of this code include: 1) The population size varies from generation to generation to save the total # of objective function evaluations. 2) The population is uniformly distributed to a number of parallel processors for simultaneous objective function evaluation. 3) The objective function evaluation can be attained from an external simulation program with control variables in its input file and objectives calculated from its output files. 4) The optimizer includes an adaptive unified differential evolution algorithm and a real value genetic algorithm. The parameters in the unified differential evolution algorithm can be chosen to attain any mutation schemes in the published literature.

Qiang, Ji↗

Heterogeneous techniques for rescaling energy deposits in the CMS Phase-2 endcap calorimeter

We present the porting to heterogeneous architectures of the algorithm used for applying linear transformations of raw energy deposits in the CMS High Granularity Calorimeter (HGCAL). This is the first heterogeneous algorithm to be fully integrated with HGCAL’s reconstruction chain. After introducing the latter and giving a brief description of the structural components of HGCAL relevant for this work, the role of the linear transformations in the calibration is reviewed. The many ways in which parallelization is achieved are described, and the successful validation of the heterogeneous algorithm is covered. Detailed performance measurements are presented, including throughput and execution time for both CPU and GPU algorithms, therefore establishing the corresponding speedup. We finally discuss the interplay between this work and the porting of other algorithms in the existing reconstruction chain, as well as integrating algorithms previously ported but not yet integrated.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Grand Unification of Quantum Algorithms

Quantum algorithms offer significant speed-ups over their classical counterparts for a variety of problems. The strongest arguments for this advantage are borne by algorithms for quantum search, quantum phase estimation, and Hamiltonian simulation, which appear as subroutines for large families of composite quantum algorithms. A number of these quantum algorithms have recently been tied together by a novel technique known as the quantum singular value transformation (QSVT), which enables one to perform a polynomial transformation of the singular values of a linear operator embedded in a unitary matrix. In the seminal GSLW’19 paper on the QSVT [Gilyén et al., ACM STOC 2019], many algorithms are encompassed, including amplitude amplification, methods for the quantum linear systems problem, and quantum simulation. Here, we provide a pedagogical tutorial through these developments, first illustrating how quantum signal processing may be generalized to the quantum eigenvalue transform, from which the QSVT naturally emerges. Paralleling GSLW’19, we then employ the QSVT to construct intuitive quantum algorithms for search, phase estimation, and Hamiltonian simulation, and also showcase algorithms for the eigenvalue threshold problem and matrix inversion. This overview illustrates how the QSVT is a single framework comprising the three major quantum algorithms, suggesting a grand unification of quantum algorithms.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Reconstruction from Uniformly Attenuated SPECT Projection Data Using the DBH Method

An algorithm was developed for the two-dimensional (2D) reconstruction of truncated and non-truncated uniformly attenuated data acquired from single photon emission computed tomography (SPECT). The algorithm is able to reconstruct data from half-scan (180o) and short-scan (180?+fan angle) acquisitions for parallel- and fan-beam geometries, respectively, as well as data from full-scan (360o) acquisitions. The algorithm is a derivative, backprojection, and Hilbert transform (DBH) method, which involves the backprojection of differentiated projection data followed by an inversion of the finite weighted Hilbert transform. The kernel of the inverse weighted Hilbert transform is solved numerically using matrix inversion. Numerical simulations confirm that the DBH method provides accurate reconstructions from half-scan and short-scan data, even when there is truncation. However, as the attenuation increases, finer data sampling is required.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Parallel Infrastructure Modeling and Inversion Module for E4D

Electrical resistivity tomography ERT is a method of imaging the electrical conductivity of the subsurface. Electrical conductivity is a useful metric for understanding the subsurface because it is governed by geomechanical and geochemical properties that drive subsurface systems. ERT works by injecting current into the subsurface across a pair of electrodes, and measuring the corresponding electrical potential response across another pair of electrodes. Many such measurements are strategically taken across an array of electrodes to produce an ERT data set. These data are then processed through a computationally demanding process known as inversion to produce an image of the subsurface conductivity structure that gave rise to the measurements. Data can be inverted to provide 2D images, 3D images, or in the case of time-lapse 3D imaging, 4D images. ERT is generally not well suited for environments with buried electrically conductive infrastructure such as pipes, tanks, or well casings, because these features tend to dominate and degrade ERT images. This reduces or eliminates the utility of ERT imaging where it would otherwise be highly useful for, for example, imaging fluid migration from leaking pipes, imaging soil contamination beneath leaking subsurface tanks, and monitoring contaminant migration in locations with dense network of metal cased monitoring wells. The location and dimension of buried metallic infrastructure is often known. If so, then the effects of the infrastructure can be explicitly modeled within the ERT imaging algorithm, and thereby removed from the corresponding ERT image. However, there are a number of obstacles limiting this application. 1) Metallic infrastructure cannot be accurately modeled with standard codes because of the large contrast in conductivity between the metal and host material. 2) Modeling infrastructure in true dimension requires the computational mesh to be highly refined near the metal inclusions, which increases computational demands. 3) The ERT imaging algorithm requires specialized modifications to accommodate high conductivity inclusions within the computational mesh. The solution to each of these challenges was implemented within E4D (formerly FERM3D), which is a parallel ERT imaging code developed at PNNL (IPID #30249). The infrastructure modeling module implement in E4D uses a method of decoupling the model at the metallic interface(s) boundaries, into several well posed sub-problems (one for each distinct metallic inclusion) that are subsequently solved and recombined to form the global solution. The approach is based on the immersed interface method, with has been applied for similar problems in other fields (e.g. semiconductor industry). Comparisons to analytic solutions have shown the results to be very accurate, addressing item 1 above. The solution is implemented about an unstructured mesh, which enables arbitrary shapes to be efficiently modelled, thereby addressing item 2 above. In addition, the algorithm is written in parallel and shows excellent scalability, which also addresses equation 2 above. Finally, because only the boundaries of metallic inclusions are modeled, there are no high conductivity cells within the modeling mesh, and the problem described by item 3 above is no longer applicable.

Johnson, TimC.↗

Acceleration of Radiance for Lighting Simulation by Using Parallel Computing with OpenCL

We report on the acceleration of annual daylighting simulations for fenestration systems in the Radiance ray-tracing program. The algorithm was optimized to reduce both the redundant data input/output operations and the floating-point operations. To further accelerate the simulation speed, the calculation for matrix multiplications was implemented using parallel computing on a graphics processing unit. We used OpenCL, which is a cross-platform parallel programming language. Numerical experiments show that the combination of the above measures can speed up the annual daylighting simulations 101.7 times or 28.6 times when the sky vector has 146 or 2306 elements, respectively.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

High-Purity Germanium Spectroscopy at Rates in Excess of 10^{6} Events/s

Abstract—In gamma spectroscopy, a compromise must be made between energy resolution and event-rate capability. Some foreseen nuclear material safeguards applications require a spectrometer with energy resolution typical of high purity germanium (HPGe) detectors, operated at rates up to and exceeding 106 events per second. We report the performance of an HPGe spectrometer adapted to run at such rates. Our system consists of a commercial semi-coaxial HPGe detector, a modified high-voltagerail, resistive-feedback, charge-sensitive preamplifier and a continuous waveform digitizer. Digitized waveforms are analyzed offline with a novel time-variant trapezoidal filter algorithm. Several time-invariant trapezoidal filters are run in parallel and the slowest one not rejected by instantaneous pileup conditions is used to measure each pulse height. We have attained full-widthat- half-maximum energy resolution of less than 8 keV measured at 662 keV with 1:08*106 per second incoming event rate and 38% throughput. An additional constraint on the width of the fast trigger filter removes a significant amount of edge pileup that passes the first pileup cut, reducing throughput to 26%. While better resolution has been reported by other authors, our throughput is over an order of magnitude higher than any other reported HPGe system operated at such an event rate.

germanium, semiconductor radiation detectors, gamm↗