Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel computing 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 1,387 records · Page 77

Energy scaling advantages of resistive memory crossbar based computation and its application to sparse coding

In this study, the exponential increase in data over the last decade presents a significant challenge to analytics efforts that seek to process and interpret such data for various applications. Neural-inspired computing approaches are being developed in order to leverage the computational properties of the analog, low-power data processing observed in biological systems. Analog resistive memory crossbars can perform a parallel read or a vector-matrix multiplication as well as a parallel write or a rank-1 update with high computational efficiency. For an N × N crossbar, these two kernels can be O(N) more energy efficient than a conventional digital memory-based architecture. If the read operation is noise limited, the energy to read a column can be independent of the crossbar size (O(1)). These two kernels form the basis of many neuromorphic algorithms such as image, text, and speech recognition. For instance, these kernels can be applied to a neural sparse coding algorithm to give an O(N) reduction in energy for the entire algorithm when run with finite precision. Sparse coding is a rich problem with a host of applications including computer vision, object tracking, and more generally unsupervised learning.

97 MATHEMATICS AND COMPUTING↗

Analog system for computing sparse codes

A parallel dynamical system for computing sparse representations of data, i.e., where the data can be fully represented in terms of a small number of non-zero code elements, and for reconstructing compressively sensed images. The system is based on the principles of thresholding and local competition that solves a family of sparse approximation problems corresponding to various sparsity metrics. The system utilizes Locally Competitive Algorithms (LCAs), nodes in a population continually compete with neighboring units using (usually one-way) lateral inhibition to calculate coefficients representing an input in an over complete dictionary.

Rozell, Christopher John↗

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↗

Communication-Avoiding and Memory-Constrained Sparse Matrix-Matrix Multiplication at Extreme Scale

Sparse matrix-matrix multiplication (SpGEMM) is a widely used kernel in various graph, scientific computing and machine learning algorithms. In this paper, we consider SpGEMMs performed on hundreds of thousands of processors generating trillions of nonzeros in the output matrix. Distributed SpGEMM at this extreme scale faces two key challenges: (1) high communication cost and (2) inadequate memory to generate the output. Furthermore, we address these challenges with an integrated communication-avoiding and memory-constrained SpGEMM algorithm that scales to 262,144 cores (more than 1 million hardware threads) and can multiply sparse matrices of any size as long as inputs and a fraction of output fit in the aggregated memory. As we go from 16,384 cores to 262,144 cores on a Cray XC40 supercomputer, the new SpGEMM algorithm runs 10x faster when multiplying large-scale protein-similarity matrices.

97 MATHEMATICS AND COMPUTING↗

Efficient Heterogeneous Execution on Large Multicore and Accelerator Platforms: Case Study Using a Block Tridiagonal Solver

The algorithmic and implementation principles are explored in gainfully exploiting GPU accelerators in conjunction with multicore processors on high-end systems with large numbers of compute nodes, and evaluated in an implementation of a scalable block tridiagonal solver. The accelerator of each compute node is exploited in combination with multicore processors of that node in performing block-level linear algebra operations in the overall, distributed solver algorithm. Optimizations incorporated include: (1) an efficient memory mapping and synchronization interface to minimize data movement, (2) multi-process sharing of the accelerator within a node to obtain balanced load with multicore processors, and (3) an automatic memory management system to efficiently utilize accelerator memory when sub-matrices spill over the limits of device memory. Results are reported from our novel implementation that uses MAGMA and CUBLAS accelerator software systems simultaneously with ACML for multithreaded execution on processors. Overall, using 940 nVidia Tesla X2090 accelerators and 15,040 cores, the best heterogeneous execution delivers a 10.9-fold reduction in run time relative to an already efficient parallel multicore-only baseline implementation that is highly optimized with intra-node and inter-node concurrency and computation-communication overlap. Detailed quantitative results are presented to explain all critical runtime components contributing to hybrid performance.

Park, Alfred J↗

2-Group global symmetries and anomalies in six-dimensional quantum field theories

We examine six-dimensional quantum field theories through the lens of higher-form global symmetries. Every Yang-Mills gauge theory in six dimensions, with field strength ƒ (2) , naturally gives rise to a continuous 1-form global symmetry associated with the 2-form instanton current J (2) ~ * Tr (ƒ (2) $\wedge$ ƒ (2) ). We show that suitable mixed anomalies involving the gauge field ƒ (2) and ordinary 0-form global symmetries, such as flavor or Poincaré symmetries, lead to continuous 2-group global symmetries, which allow two flavor currents or two stress tensors to fuse into the 2-form current J (2) . We discuss several features of 2-group symmetry in six dimensions, many of which parallel the four-dimensional case. The majority of six-dimensional supersymmetric conformal field theories (SCFTs) and little string theories have infrared phases with non-abelian gauge fields. We show that the mixed anomalies leading to 2-group symmetries can be present in little string theories, but that they are necessarily absent in SCFTs. This allows us to establish a previously conjectured algorithm for computing the ’t Hooft anomalies of most SCFTs from the spectrum of weakly-coupled massless particles on the tensor branch of these theories. We then apply this understanding to prove that the a -type Weyl anomaly of all SCFTs with a tensor branch must be positive, a > 0.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

AMReX

AMReX is a software framework for the development and implementation of massively parallel algorithms using block-structured adaptive mesh refinement (AMR). Block-structured AMR provides the basis for the temporal and spatial discretization strategy for a large variety of applications of interest to the DOE. AMReX provides a unified infrastructure with the functionality needed for applications to be able to effectively and efficiently utilize machines from laptops to exascale accelerator-based architectures. AMR reduces the computational cost and memory footprint compared to a uniform mesh while preserving accurate descriptions of different physical processes in complex multi-physics algorithms. AMReX supports algorithms that solve systems of partial differential equations (PDEs) in simple or complex geometries, and those that use particles and/or particle-mesh operations to represent component physical processes.

Zhang, Weiqun↗

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↗

Hierarchical computation in the SPMD programming model

Large-scale parallel machines are programmed mainly with the single program, multiple data (SPMD) model of parallelism. While this model has advantages of scalability and simplicity, it does not fit well with divide-and-conquer parallelism or hierarchical machines that mix shared and distributed memory. In this paper, we define the recursive single program, multiple data model (RSPMD) that extends SPMD with a hierarchical team mechanism to support hierarchical algorithms and machines. We implement this model in the Titanium language and describe how to eliminate a class of deadlocks by ensuring alignment of collective operations. We present application case studies evaluating the RSPMD model, showing that it enables divide-and-conquer algorithms such as sorting to be elegantly expressed and that team collective operations increase performance of conjugate gradient by up to a factor of two. The model also facilitates optimizations for hierarchical machines, improving scalability of particle in cell by 8x and performance of sorting and a stencil code by up to 40% and 14%, respectively.

Kamil, A↗

Final Report: Towards Optimal Petascale Simulations (TOPS), ER25785

Multiscale, multirate scientific and engineering applications in the SciDAC portfolio possess resolution requirements that are practically inexhaustible and demand execution on the highest-capability computers available, which will soon reach the petascale. While the variety of applications is enormous, their needs for mathematical software infrastructure are surprisingly coincident; moreover the chief bottleneck is often the solver. At their current scalability limits, many applications spend a vast majority of their operations in solvers, due to solver algorithmic complexity that is superlinear in the problem size, whereas other phases scale linearly. Furthermore, the solver may be the phase of the simulation with the poorest parallel scalability, due to intrinsic global dependencies. This project brings together the providers of some of the world's most widely distributed, freely available, scalable solver software and focuses them on relieving this bottleneck for many specific applications within SciDAC, which are representative of many others outside. Solver software directly supported under TOPS includes: hypre, PETSc, SUNDIALS, SuperLU, TAO, and Trilinos. Transparent access is also provided to other solver software through the TOPS interface. The primary goals of TOPS are the development, testing, and dissemination of solver software, especially for systems governed by PDEs. Upon discretization, these systems possess mathematical structure that must be exploited for optimal scalability; therefore, application-targeted algorithmic research is included. TOPS software development includes attention to high performance as well as interoperability among the solver components. Support for integration of TOPS solvers into SciDAC applications is also directly supported by this proposal. The role of the UCSD PI in this overall CET, is one of direct interaction between the TOPS software partners and various DOE applications scientists' specifically toward magnetohydrodynamics (MHD) simulations with the Center for Extended Magnetohydrodynamic Modeling (CEMM) SciDAC and Applied Partial Differential Equations Center (APDEC) SciDAC, and toward core-collapse supernova simulations with the previous Terascale Supernova Initiative (TSI) SciDAC and in continued work on INCITE projects headed by Doug Swesty, SUNY Stony Brook. In addition to these DOE applications scientists, the UCSD PI works to bring leading-edge DOE solver technology to applications scientists in cosmology and large-scale galactic structure formation. Unfortunately, the funding for this grant ended after only two years of its five-year duration, in August 2008, due to difficulties at DOE in transferring the grant to the PI's new faculty position at Southern Methodist University. Therefore, this report only describes two years' worth of effort.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Discovering causal structure with reproducing-kernel Hilbert space ε -machines

We merge computational mechanics’ definition of causal states (predictively equivalent histories) with reproducing-kernel Hilbert space (RKHS) representation inference. The result is a widely applicable method that infers causal structure directly from observations of a system’s behaviors whether they are over discrete or continuous events or time. A structural representation—a finite- or infinite-state kernel ϵ-machine—is extracted by a reduced-dimension transform that gives an efficient representation of causal states and their topology. In this way, the system dynamics are represented by a stochastic (ordinary or partial) differential equation that acts on causal states. We introduce an algorithm to estimate the associated evolution operator. Paralleling the Fokker–Planck equation, it efficiently evolves causal-state distributions and makes predictions in the original data space via an RKHS functional mapping. We demonstrate these techniques, together with their predictive abilities, on discrete-time, discrete-value infinite Markov-order processes generated by finite-state hidden Markov models with (i) finite or (ii) uncountably infinite causal states and (iii) continuous-time, continuous-value processes generated by thermally driven chaotic flows. The method robustly estimates causal structure in the presence of varying external and measurement noise levels and for very high-dimensional data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

On using moving windows in finite element time domain simulation for long accelerator structures

A finite element moving window technique is developed to simulate the propagation of electromagnetic waves induced by the transit of a charged particle beam inside large and long structures. The window moving along with the beam in the computational domain adopts high-order finite element basis functions through p refinement and/or a high-resolution mesh through h refinement so that a sufficient accuracy is attained with substantially reduced computational costs. Algorithms to transfer discretized fields from one mesh to another, which are the keys to implementing a moving window in a finite element unstructured mesh, are presented. Numerical experiments are carried out using the moving window technique to compute short-range wakefields in long accelerator structures. The results are compared with those obtained from the normal finite element time domain (FETD) method and the advantages of using the moving window technique are discussed.

97 MATHEMATICS AND COMPUTING↗

Distributed non-equilibrium Green’s function algorithms for the simulation of nanoelectronic devices with scattering

Through the non-equilibrium Green’s function (NEGF) formalism, quantum-scale device simulation can be performed with the inclusion of electron-phonon scattering. Yet, the simulation of realistically sized devices under the NEGF formalism typically requires prohibitive amounts of memory and computation time. Two of the most demanding computational problems for NEGF simulation involve mathematical operations with structured matrices called semiseparable matrices. In this work, we present parallel approaches for these computational problems which allow for efficient distribution of both memory and computation based upon the underlying device structure. This is vital when simulating realistically sized devices due to the aforementioned computational burdens. First, we consider determining a distributed compact representation for the retarded Green’s function matrix G R . This compact representation is exact and allows for any entry in the matrix to be generated through the inherent semiseparable structure. The second parallel operation allows for the computation of electron density and current characteristics for the device. Specifically, matrix products between the distributed representation for the semiseparable matrix G R and the self-energy scattering terms in Σ < produce the less-than Green’s function G < . As an illustration of the computational efficiency of our approach, we stably generate the mobility for nanowires with cross-sectional sizes as large as 4.5 nm, assuming an atomistic model with scattering.

47 OTHER INSTRUMENTATION↗

Fast analysis of molecular dynamics trajectories with graphics processing units-Radial distribution function histogramming

The calculation of radial distribution functions (RDFs) from molecular dynamics trajectory data is a common and computationally expensive analysis task. The rate limiting step in the calculation of the RDF is building a histogram of the distance between atom pairs in each trajectory frame. Here we present an implementation of this histogramming scheme for multiple graphics processing units (GPUs). The algorithm features a tiling scheme to maximize the reuse of data at the fastest levels of the GPU's memory hierarchy and dynamic load balancing to allow high performance on heterogeneous configurations of GPUs. Several versions of the RDF algorithm are presented, utilizing the specific hardware features found on different generations of GPUs. We take advantage of larger shared memory and atomic memory operations available on state-of-the-art GPUs to accelerate the code significantly. The use of atomic memory operations allows the fast, limited-capacity on-chip memory to be used much more efficiently, resulting in a fivefold increase in performance compared to the version of the algorithm without atomic operations. The ultimate version of the algorithm running in parallel on four NVIDIA GeForce GTX 480 (Fermi) GPUs was found to be 92 times faster than a multithreaded implementation running on an Intel Xeon 5550 CPU. On this multi-GPU hardware, the RDF between two selections of 1,000,000 atoms each can be calculated in 26.9 s per frame. The multi-GPU RDF algorithms described here are implemented in VMD, a widely used and freely available software package for molecular dynamics visualization and analysis.

97 MATHEMATICS AND COMPUTING↗

Status on the Development of a Modeling and Simulation Framework for the Economic Assessment of Nuclear Hybrid Energy Systems

An effort to design and build a modeling and simulation framework to assess the economic viability of Nuclear Hybrid Energy Systems (NHES) was undertaken in fiscal year 2015 (FY15). The purpose of this report is to document the various tasks associated with the development of such a framework and to provide a status on its progress. Several tasks have been accomplished. First, starting from a simulation strategy, a rigorous mathematical formulation has been achieved in which the economic optimization of a Nuclear Hybrid Energy System is presented as a constrained robust (under uncertainty) optimization problem. Some possible algorithms for the solution of the optimization problem are presented. A variation of the Simultaneous Perturbation Stochastic Approximation algorithm has been implemented in RAVEN and preliminary tests have been performed. The development of the software infrastructure to support the simulation of the whole NHES has also moved forward. The coupling between RAVEN and an implementation of the Modelica language (OpenModelica) has been implemented, migrated under several operating systems and tested using an adapted model of a desalination plant. In particular, this exercise was focused on testing the coupling of the different code systems; testing parallel, computationally expensive simulations on the INL cluster; and providing a proof of concept for the possibility of using surrogate models to represent the different NHES subsystems. Another important step was the porting of the RAVEN code under the Windows™ operating system. This accomplishment makes RAVEN compatible with the development environment that is being used for dynamic simulation of NHES components. A very simplified model of a NHES on the electric market has been built in RAVEN to confirm expectations on the analysis capability of RAVEN to provide insight into system economics and to test the capability of RAVEN to identify limit surfaces even for stochastic constraints. This capability will be needed in the future to enforce the stochastic constraints on the electric demand coverage from the NHES. The development team gained experience with many of the tools that are currently envisioned for use in the economic analysis of NHES and completed several important steps. Given the complexity of the project, preference has been given to a structural approach in which several independent efforts have been used to build the cornerstone of the simulation framework. While this is good approach in establishing such a complex framework, it may delay reaching more complete results on the performance of analyzed system configurations. The integration of the previously reported exergy analysis approach was initially proposed as part of this milestone. However, in reality, the exergy-based apportioning of cost will take place only in a second stage of the implementation since it will be used to properly allocate cost among the different NHES subsystems. Therefore, exergy does not appear at the level of the main drivers in the analysis framework; the latter development of the base framework is the focus of this report.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Material Science Image Analysis using Quant-CT in ImageJ

We introduce a computational analysis workflow to access properties of solid objects using nondestructive imaging techniques that rely on X-ray imaging. The goal is to process and quantify structures from material science sample cross sections. The algorithms can differentiate the porous media (high density material) from the void (background, low density media) using a Boolean classifier, so that we can extract features, such as volume, surface area, granularity spectrum, porosity, among others. Our workflow, Quant-CT, leverages several algorithms from ImageJ, such as statistical region merging and 3D object counter. It also includes schemes for bilateral filtering that use a 3D kernel, for parallel processing of sub-stacks, and for handling over-segmentation using histogram similarities. The Quant-CT supports fast user interaction, providing the ability for the user to train the algorithm via subsamples to feed its core algorithms with automated parameterization. Quant-CT plugin is currently available for testing by personnel at the Advanced Light Source and Earth Sciences Divisions and Energy Frontier Research Center (EFRC), LBNL, as part of their research on porous materials. The goal is to understand the processes in fluid-rock systems for the geologic sequestration of CO2, and to develop technology for the safe storage of CO2 in deep subsurface rock formations. We describe our implementation, and demonstrate our plugin on porous material images. This paper targets end-users, with relevant information for developers to extend its current capabilities.

97 MATHEMATICS AND COMPUTING↗

Big Data Seismology

The discipline of seismology is based on observations of ground motion that are inherently undersampled in space and time. Our basic understanding of earthquake processes and our ability to resolve 4D Earth structure are fundamentally limited by data volume. Currently, Big Data Seismology is an emergent revolution involving the use of large, data-dense inquiries that is providing new opportunities to make fundamental advances in these areas. This article reviews recent scientific advances enabled by Big Data Seismology through the context of three major drivers: the development of new data-dense sensor systems, improvements in computing, and the development of new types of techniques and algorithms. Each driver is explored in the context of both global and exploration seismology, alongside collaborative opportunities that combine the features of long-duration data collections (common to global seismology) with dense networks of sensors (common to exploration seismology). The review explores some of the unique challenges and opportunities that Big Data Seismology presents, drawing on parallels from other fields facing similar issues. Finally, recent scientific findings enabled by dense seismic data sets are discussed, and we assess the opportunities for significant advances made possible with Big Data Seismology. This review is designed to be a primer for seismologists who are interested in getting up-to-speed with how the Big Data revolution is advancing the field of seismology.

58 GEOSCIENCES↗

Scalable, Multithreaded, Partially-in-Place Sorting

A recent trend in hardware development is producing computing systems that are stretching the number of cores and size of shared-memory beyond where most fundamental serial algorithms perform well. The expectation is that this trend will continue. So it makes sense to rethink our fundamental algorithms such as sorting. There are many situations where data that needs to be sorted will actually fit into the shared memory so applications could benefit from an efficient parallel sorting algorithm. When sorting large data (at least hundreds of Gigabytes) in a single shared memory, there are two factors that affect the algorithm choice. First, does the algorithm sort in-place? And second, does the algorithm scale well beyond tens of threads? Surprisingly, existing algorithms posses either one of these factors, but not both. We present an approach that gracefully degrades in performance as the amount of available working memory decreases relative to the size of the input.

HPC↗