Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel algorithm”

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 1,639 records · Page 91

Work stealing for GPU-accelerated parallel programs in a global address space framework

Task parallelism is an attractive approach to automatically load balance the computation in a parallel system and adapt to dynamism exhibited by parallel systems. Exploiting task parallelism through work stealing has been extensively studied in shared and distributed-memory contexts. In this paper, we study the design of a system that uses work stealing for dynamic load balancing of task-parallel programs executed on hybrid distributed-memory CPU-graphics processing unit (GPU) systems in a global-address space framework. We take into account the unique nature of the accelerator model employed by GPUs, the significant performance difference between GPU and CPU execution as a function of problem size, and the distinct CPU and GPU memory domains. We consider various alternatives in designing a distributed work stealing algorithm for CPU-GPU systems, while taking into account the impact of task distribution and data movement overheads. These strategies are evaluated using microbenchmarks that capture various execution configurations as well as the state-of-the-art CCSD(T) application module from the computational chemistry domain

97 MATHEMATICS AND COMPUTING↗

Continuation and bifurcation analysis of large-scale dynamical systems with LOCA.

Dynamical systems theory provides a powerful framework for understanding the behavior of complex evolving systems. However applying these ideas to large-scale dynamical systems such as discretizations of multi-dimensional PDEs is challenging. Such systems can easily give rise to problems with billions of dynamical variables, requiring specialized numerical algorithms implemented on high performance computing architectures with thousands of processors. This talk will describe LOCA, the Library of Continuation Algorithms, a suite of scalable continuation and bifurcation tools optimized for these types of systems that is part of the Trilinos software collection. In particular, we will describe continuation and bifurcation analysis techniques designed for large-scale dynamical systems that are based on specialized parallel linear algebra methods for solving augmented linear systems. We will also discuss several other Trilinos tools providing nonlinear solvers (NOX), eigensolvers (Anasazi), iterative linear solvers (AztecOO and Belos), preconditioners (Ifpack, ML, Amesos) and parallel linear algebra data structures (Epetra and Tpetra) that LOCA can leverage for efficient and scalable analysis of large-scale dynamical systems.

99 GENERAL AND MISCELLANEOUS↗

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↗

Optimizing checkpoint data placement with guaranteed burst buffer endurance in large-scale hierarchical storage systems

Non-volatile devices, such as SSDs, will be an integral part of the deepening storage hierarchy on large-scale HPC systems. These devices can be on the compute nodes as part of a distributed burst buffer service or they can be external. Wherever they are located in the hierarchy, one critical design issue is the SSD endurance under the write-heavy workloads, such as the checkpoint I/O for scientific applications. For these environments, it is widely assumed that checkpoint operations can occur once every 60 min and for each checkpoint step as much as half of the system memory can be written out. Unfortunately, for large-scale HPC applications, the burst buffer SSDs can be worn out much more quickly given the extensive amount of data written at every checkpoint step. One possible solution is to control the amount of data written by reducing the checkpoint frequency. However, a direct effect caused by reduced checkpoint frequency is the increased vulnerability window of system failures and therefore potentially wasted computation time, especially for large-scale compute jobs.

Computer Science↗

Fluorescence Color by Data-Driven Design of Genomic Silver Clusters

DNA nucleobase sequence controls the size of DNA-stabilized silver clusters, leading to their well-known yet little understood sequence-tuned colors. The enormous space of possible DNA sequences for templating clusters has challenged the understanding of how sequence selects cluster properties and has limited the design of applications that employ these clusters. Here, we investigate the genomic role of DNA sequence for fluorescent silver clusters using a data-driven approach. Employing rapid parallel silver cluster synthesis and fluorimetry, we determine the fluorescence spectra of silver cluster products stabilized by 1432 distinct DNA oligomers. By applying pattern recognition algorithms to this large experimental data set, we discover certain DNA base patterns, or “motifs,” that correlate to silver clusters with similar fluorescence spectra. These motifs are employed in machine learning classifiers to predictively design DNA template sequences for specific fluorescence color bands. Our method improves selectivity of templates by 330% for silver clusters with peak emission wavelengths beyond 660 nm. The discovered base motifs also provide physical insights into how DNA sequence controls silver cluster size and color. Finally, this predictive design approach for color of DNA-stabilized silver clusters exhibits the potential of machine learning and data mining to increase the precision and efficiency of nanomaterials design, even for a soft-matter-inorganic hybrid system characterized by an extremely large parameter space.

59 BASIC BIOLOGICAL SCIENCES↗

Progress in modelling fast-ion D-alpha spectra and neutral particle analyzer fluxes using FIDASIM

FIDASIM is a code that models signals produced by charge-exchange reactions between neutrals and ions (both fast and thermal) in magnetically confined plasmas. With the ion distribution function as input, the code predicts the efflux to a neutral particle analyzer diagnostic and the photon radiance of Balmer-alpha light to a fast-ion D α diagnostic, in addition to many other related quantities. A new, parallelized version of the Monte Carlo code FIDASIM has been developed in Fortran90 that is substantially faster than the original interactive data language version. Modified algorithms include more accurate treatments of the time dependent collisional-radiative equations that describe neutral energy levels, of the cloud of ‘halo’ neutrals that surround the injected neutral beam, and of finite Larmor radius effects. Enhanced physics capabilities include modelling ‘passive’ signals from cold edge neutrals, the ability to treat general three-dimensional magnetic confinement configurations, and calculations of diagnostic-specific weight functions that enable tomographic reconstructions of the fast-ion distribution function. Neutral beam attenuation, beam emission, and fast-ion birth profiles are also modelled. Finally, the new algorithms have been successfully validated against experimental data and new features have been tested through benchmarks between two independently developed versions of the code.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

What does fault tolerant Deep Learning need from MPI?

Deep Learning (DL) algorithms have become the {\em de facto} Machine Learning (ML) algorithm for large scale data analysis. DL algorithms are computationally expensive -- even distributed DL implementations which use MPI require days of training (model learning) time on commonly studied datasets. Long running DL applications become susceptible to faults -- requiring development of a fault tolerant system infrastructure, in addition to fault tolerant DL algorithms. This raises an important question: {\em What is needed from MPI for designing fault tolerant DL implementations?} In this paper, we address this problem for permanent faults. We motivate the need for a fault tolerant MPI specification by an in-depth consideration of recent innovations in DL algorithms and their properties, which drive the need for specific fault tolerance features. We present an in-depth discussion on the suitability of different parallelism types (model, data and hybrid); a need (or lack thereof) for check-pointing of any critical data structures; and most importantly, consideration for several fault tolerance proposals (user-level fault mitigation (ULFM), Reinit) in MPI and their applicability to fault tolerant DL implementations. We leverage a distributed memory implementation of Caffe, currently available under the Machine Learning Toolkit for Extreme Scale (MaTEx). We implement our approaches by extending MaTEx-Caffe for using ULFM-based implementation. Our evaluation using the ImageNet dataset and AlexNet neural network topology demonstrates the effectiveness of the proposed fault tolerant DL implementation using OpenMPI based ULFM.

Amatya, Vinay C.↗

Dictionary Learning with Accumulator Neurons

The Locally Competitive Algorithm (LCA) uses local competition between non-spiking leaky integrator neurons to infer sparse representations, allowing for potentially real-time execution on massively parallel neuromorphic architectures such as Intel's Loihi processor. Here, we focus on the problem of inferring sparse representations from streaming video using dictionaries of spatiotemporal features optimized in an unsupervised manner for sparse reconstruction. Non-spiking LCA has previously been used to achieve unsupervised learning of spatiotemporal dictionaries composed of convolutional kernels from raw, unlabeled video. We demonstrate how unsupervised dictionary learning with spiking LCA (\hbox{S-LCA}) can be efficiently implemented using accumulator neurons, which combine a conventional leaky-integrate-and-fire (\hbox{LIF}) spike generator with an additional state variable that is used to minimize the difference between the integrated input and the spiking output. We demonstrate dictionary learning across a wide range of dynamical regimes, from graded to intermittent spiking, for inferring sparse representations of both static images drawn from the CIFAR database as well as video frames captured from a DVS camera. On a classification task that requires identification of the suite from a deck of cards being rapidly flipped through as viewed by a DVS camera, we find essentially no degradation in performance as the LCA model used to infer sparse spatiotemporal representations migrates from graded to spiking. We conclude that accumulator neurons are likely to provide a powerful enabling component of future neuromorphic hardware for implementing online unsupervised learning of spatiotemporal dictionaries optimized for sparse reconstruction of streaming video from event based DVS cameras.

artificial intelligence↗

PARTISN on Advanced/Heterogeneous Processing Systems [Slides]

This presentation describes both the importance and the difficulty of parallelizing deterministic neutron transport calculations using the LANL code PARTISN. A brief history of the standard "KBA" parallelization method used for deterministic SN transport is presented, then the modifications necessary to implement this on Roadrunner are covered. The use of the Roadrunner algorithm with OpenMP is then described, along with a brief mention of an alternative thermal radiation transport algorithm for GPUs. Finally, ongoing work with PARTISN on the MIC many-core architecture is noted.

97 MATHEMATICS AND COMPUTING↗

High Fidelity Simulations of Large-scale Wireless Networks (Part II - FY2017)

The ability to simulate wireless networks at large-scale for meaningful amount of time is considerably lacking in today's network simulators. For this reason, many published work in this area often limit their simulation studies to less than a 1,000 nodes and either over-simplify channel characteristics or perform studies over time scales much less than a day. In this report, we show that one can overcome these limitations and study problems of high practical consequence. This work presents two key contributions to high fidelity simulation of large-scale wireless networks: (a) wireless simulations can be sped up by more than 100X in runtime using ideas from spatial indexing algorithms and clipping of negligible signals and (b) clustering and task-oriented programming paradigm can be used to reduce inter- process communication in a parallel discrete event simulation resulting in a better scaling efficiency.

97 MATHEMATICS AND COMPUTING↗

Data Access Performance Through Parallelization and Vectored Access: Some Results

High Energy Physics data processing and analysis applications typically deal with the problem of accessing and processing data at high speed. Recent studies, development and test work have shown that the latencies due to data access can often be hidden by parallelizing them with the data processing, thus giving the ability to have applications which process remote data with a high level of efficiency. Techniques and algorithms able to reach this result have been implemented in the client side of the Scalla/xrootd system, and in this contribution we describe the results of some tests done in order to compare their performance and characteristics. These techniques, if used together with multiple streams data access, can also be effective in allowing to efficiently and transparently deal with data repositories accessible via a Wide Area Network.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Dakota A Multilevel Parallel Object-Oriented Framework for Design Optimization Parameter Estimation Uncertainty Quantification and Sensitivity Analysis (V.6.14) (Theory Manual)

The Dakota toolkit provides a flexible and extensible interface between simulation codes and iterative analysis methods. Dakota contains algorithms for optimization with gradient and nongradient-based methods; uncertainty quantification with sampling, reliability, and stochastic expansion methods; parameter estimation with nonlinear least squares methods; and sensitivity/variance analysis with design of experiments and parameter study methods. These capabilities may be used on their own or as components within advanced strategies such as surrogate-based optimization, mixed integer nonlinear programming, or optimization under uncertainty. By employing object-oriented design to implement abstractions of the key components required for iterative systems analyses, the Dakota toolkit provides a flexible and extensible problem-solving environment for design and performance analysis of computational models on high performance computers. This report serves as a theoretical manual for selected algorithms implemented within the Dakota software. It is not intended as a comprehensive theoretical treatment, since a number of existing texts cover general optimization theory, statistical analysis, and other introductory topics. Rather, this manual is intended to summarize a set of Dakota-related research publications in the areas of surrogate-based optimization, uncertainty quantification, and optimization under uncertainty that provide the foundation for many of Dakota's iterative analysis capabilities.

97 MATHEMATICS AND COMPUTING↗

Dakota, A Multilevel Parallel Object-Oriented Framework for Design Optimization, Parameter Estimation, Uncertainty Quantification, and Sensitivity Analysis: Theory Manual (V.6.15)

The Dakota toolkit provides a flexible and extensible interface between simulation codes and iterative analysis methods. Dakota contains algorithms for optimization with gradient and nongradient-based methods; uncertainty quantification with sampling, reliability, and stochastic expansion methods; parameter estimation with nonlinear least squares methods; and sensitivity/variance analysis with design of experiments and parameter study methods. These capabilities may be used on their own or as components within advanced strategies such as surrogate-based optimization, mixed integer nonlinear programming, or optimization under uncertainty. By employing object-oriented design to implement abstractions of the key components required for iterative systems analyses, the Dakota toolkit provides a flexible and extensible problem-solving environment for design and performance analysis of computational models on high performance computers. This report serves as a theoretical manual for selected algorithms implemented within the Dakota software. It is not intended as a comprehensive theoretical treatment, since a number of existing texts cover general optimization theory, statistical analysis, and other introductory topics. Rather, this manual is intended to summarize a set of Dakota-related research publications in the areas of surrogate-based optimization, uncertainty quantification, and optimization under uncertainty that provide the foundation for many of Dakota's iterative analysis capabilities.

97 MATHEMATICS AND COMPUTING↗

dfnWorks: A discrete fracture network framework for modeling subsurface flow and transport

DFNWORKS is a parallelized computational suite to generate three-dimensional discrete fracture networks (DFN) and simulate flow and transport. Developed at Los Alamos National Laboratory over the past five years, it has been used to study flow and transport in fractured media at scales ranging from millimeters to kilometers. The networks are created and meshed using DFNGEN, which combines FRAM (the feature rejection algorithm for meshing) methodology to stochastically generate three-dimensional DFNs with the LaGriT meshing toolbox to create a high-quality computational mesh representation. The representation produces a conforming Delaunay triangulation suitable for high performance computing finite volume solvers in an intrinsically parallel fashion. Flow through the network is simulated in dfnFlow, which utilizes the massively parallel subsurface flow and reactive transport finite volume code PFLOTRAN. A Lagrangian approach to simulating transport through the DFN is adopted within DFNTRANS to determine pathlines and solute transport through the DFN. Example applications of this suite in the areas of nuclear waste repository science, hydraulic fracturing and CO 2 sequestration are also included.

42 ENGINEERING↗

Explicit integration with GPU acceleration for large kinetic networks

We demonstrate the first implementation of recently-developed fast explicit kinetic integration algorithms on modern graphics processing unit (GPU) accelerators. Taking as a generic test case a Type Ia supernova explosion with an extremely stiff thermonuclear network having 150 isotopic species and 1604 reactions coupled to hydrodynamics using operator splitting, we demonstrate the capability to solve of order 100 realistic kinetic networks in parallel in the same time that standard implicit methods can solve a single such network on a CPU. This orders-of-magnitude decrease in computation time for solving systems of realistic kinetic networks implies that important coupled, multiphysics problems in various scientific and technical fields that were intractable, or could be simulated only with highly schematic kinetic networks, are now computationally feasible.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multigrid Reduction in Time for Chaotic Dynamical Systems

As CPU clock speeds have stagnated and high performance computers continue to have ever higher core counts, increased parallelism is needed to take advantage of these new architectures. Traditional serial time-marching schemes can be a significant bottleneck, as many types of simulations require large numbers of time-steps which must be computed sequentially. Parallel-in-time schemes, such as the Multigrid Reduction in Time (MGRIT) method, remedy this by parallelizing across time-steps and have shown promising results for parabolic problems. However, chaotic problems have proved more difficult, since chaotic initial value problems (IVPs) are inherently ill-conditioned. MGRIT relies on a hierarchy of successively coarser time-grids to iteratively correct the solution on the finest time-grid, but due to the nature of chaotic systems, small inaccuracies on the coarser levels can be greatly magnified and lead to poor coarse-grid corrections. Here we introduce a modified MGRIT algorithm based on an existing quadratically converging nonlinear extension to the multigrid Full Approximation Scheme (FAS), as well as a novel time-coarsening scheme. Together, these approaches better capture long-term chaotic behavior on coarse-grids and greatly improve convergence of MGRIT for chaotic IVPs. Further, we introduce a novel low-memory variant of the algorithm for solving chaotic PDEs with MGRIT which not only solves the IVP, but also provides estimates for the unstable Lyapunov vectors of the system. Finally, we provide supporting numerical results for the Lorenz system and demonstrate parallel speedup for the chaotic Kuramoto–Sivashinsky PDE over a significantly longer time-domain than in previous works.

97 MATHEMATICS AND COMPUTING↗

Quantum simulation of excited states from parallel contracted quantum eigensolvers

Abstract Computing excited-state properties of molecules and solids is considered one of the most important near-term applications of quantum computers. While many of the current excited-state quantum algorithms differ in circuit architecture, specific exploitation of quantum advantage, or result quality, one common feature is their rooting in the Schrödinger equation. However, through contracting (or projecting) the eigenvalue equation, more efficient strategies can be designed for near-term quantum devices. Here we demonstrate that when combined with the Rayleigh–Ritz variational principle for mixed quantum states, the ground-state contracted quantum eigensolver (CQE) can be generalized to compute any number of quantum eigenstates simultaneously. We introduce two excited-state (anti-Hermitian) CQEs that perform the excited-state calculation while inheriting many of the remarkable features of the original ground-state version of the algorithm, such as its scalability. To showcase our approach, we study several model and chemical Hamiltonians and investigate the performance of different implementations.

Physics↗