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 667 records · Page 37

Parallel transport sweeps on two-dimensional cartesian and hexagonal grids

This paper aims to provide a proof of concept for parallel transport sweeps on two-dimensional hexagonal grids for the discrete ordinates transport equation. While the method is an extension of the popular and well-established Koch-Baker-Alcoulffe (KBA) algorithm, there are significant differences between the cartesian and hexagonal grid and thereafter sweep. The most important is the three-way connectivity of hexagons within the grid which creates greater dependencies between the elements. The KBA method in structured orthogonal grids was first implemented in the DRAGON5 code and the method is first described here. The differences in implementation for the hexagonal grid are also described. Benchmark results are also presented, showing roughly 10 times speedup in computational times with roughly 100 processors, in both cases. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Symplectic multi-particle tracking on GPUs

A symplectic multi-particle tracking model is implemented on the Graphic Processing Units (GPUs) using the Compute Unified Device Architecture (CUDA) language. The symplectic tracking model can preserve phase space structure and reduce non-physical effects in long term simulation, which is important for beam property evaluation in particle accelerators. Though this model is computationally expensive, it is very suitable for parallelization and can be accelerated significantly by using GPUs. In this paper, we optimized the implementation of the symplectic tracking model on both single GPU and multiple GPUs. Using a single GPU processor, the code achieves a factor of 2-10 speedup for a range of problem sizes compared with the time on a single state-of-the-art Central Processing Unit (CPU) node with similar power consumption and semiconductor technology. It also shows good scalability on a multi-GPU cluster at Oak Ridge Leadership Computing Facility. In an application to beam dynamics simulation, the GPU implementation helps save more than a factor of two total computing time in comparison to the CPU implementation.

43 PARTICLE ACCELERATORS↗

Hard scaling challenges for ab initio molecular dynamics capabilities in NWChem: Using 100,000 CPUs per second

An overview of the parallel algorithms for ab initio molecular dynamics (AIMD) used in the NWChem program package is presented, including recent developments for computing exact exchange. These algorithms make use of a two-dimensional processor geometry proposed by Gygi et al for use in AIMD algorithms. Using this strategy, a highly scalable algorithm for exact exchange has been developed and incorporated it into AIMD. This new algorithm for exact exchange employs an incomplete butterfly to overcome the bottleneck associated with exact exchange term, and it makes judicious use of data replication. Initial testing has shown that this algorithm can scale to over 20,000 CPUs even for modest size simulation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

CMS multicore scheduling strategy

In the next years, processor architectures based on much larger numbers of cores will be most likely the model to continue 'Moore's Law' style throughput gains. This not only results in many more jobs in parallel running the LHC Run 1 era monolithic applications, but also the memory requirements of these processes push the workernode architectures to the limit. One solution is parallelizing the application itself, through forking and memory sharing or through threaded frameworks. CMS is following all of these approaches and has a comprehensive strategy to schedule multicore jobs on the GRID based on the glideinWMS submission infrastructure. The main component of the scheduling strategy, a pilot-based model with dynamic partitioning of resources that allows the transition to multicore or whole-node scheduling without disallowing the use of single-core jobs, is described. This contribution also presents the experiences made with the proposed multicore scheduling schema and gives an outlook of further developments working towards the restart of the LHC in 2015.

Perez-Calero Yzquierdo, Antonio↗

TGE: Machine Learning Based Task Graph Embedding for Large-Scale Topology Mapping

Task mapping is an important problem in parallel and distributed computing. The goal in task mapping is to find an optimal layout of the processes of an application (or a task) onto a given network topology. We target this problem in the context of staging applications. A staging application consists of two or more parallel applications (also referred to as staging tasks) which run concurrently and exchange data over the course of computation. Task mapping becomes a more challenging problem in staging applications, because not only data is exchanged between the staging tasks, but also the processes of a staging task may exchange data with each other. We propose a novel method, called Task Graph Embedding (TGE), that harnesses the observable graph structures of parallel applications and network topologies. TGE employs a machine learning based algorithm to find the best representation of a graph, called an embedding, onto a space in which the task-to-processor mapping problem can be solved. We evaluate and demonstrate the effectiveness of TGE experimentally with the communication patterns extracted from runs of XGC, a large-scale fusion simulation code, on Titan.

Choi, Jong Youl↗

A high-performance parallel algorithm for nonnegative matrix factorization

Non-negative matrix factorization (NMF) is the problem of determining two non-negative low rank factors W and H, for the given input matrix A, such that A ≈ WH. NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient distributed algorithms to solve the problem for big data sets. We propose a high-performance distributed-memory parallel algorithm that computes the factorization by iteratively solving alternating non-negative least squares (NLS) subproblems for W and H. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). As opposed to previous implementations, our algorithm is also flexible: (1) it performs well for both dense and sparse matrices, and (2) it allows the user to choose any one of the multiple algorithms for solving the updates to low rank factors W and H within the alternating iterations. We demonstrate the scalability of our algorithm and compare it with baseline implementations, showing significant performance improvements.

Kannan, Ramakrishnan↗

HPC-NMF: A High-Performance Parallel Algorithm for Nonnegative Matrix Factorization

NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient distributed algorithms to solve the problem for big data sets. We propose a high-performance distributed-memory parallel algorithm that computes the factorization by iteratively solving alternating non-negative least squares (NLS) subproblems for $\WW$ and $\HH$. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). As opposed to previous implementation, our algorithm is also flexible: It performs well for both dense and sparse matrices, and allows the user to choose any one of the multiple algorithms for solving the updates to low rank factors $\WW$ and $\HH$ within the alternating iterations.

Kannan, Ramakrishnan↗

Final Report, DE-FG01-06ER25718 Domain Decomposition and Parallel Computing

The goal of this project is to develop and improve domain decomposition algorithms for a variety of partial differential equations such as those of linear elasticity and electro-magnetics.These iterative methods are designed for massively parallel computing systems and allow the fast solution of the very large systems of algebraic equations that arise in large scale and complicated simulations. A special emphasis is placed on problems arising from Maxwell's equation. The approximate solvers, the preconditioners, are combined with the conjugate gradient method and must always include a solver of a coarse model in order to have a performance which is independent of the number of processors used in the computer simulation. A recent development allows for an adaptive construction of this coarse component of the preconditioner.

97 MATHEMATICS AND COMPUTING↗

A 2nd Generation Parallel Advancing Front Grid Generator

A scalable, parallel advancing grid generation technique has been developed for complex geometries and meshes with large size variations. The key innovation compared to previous techniques is the use of a domain-defining grid that has the same fine surface triangulation as the final mesh desired, but a much coarser interior mesh. In this way, the domain to be gridded is uniquely defined, overcoming a shortcoming of previous approaches. This domain-defining grid is then partitioned according to the estimated number of elements to be generated, allowing for a balanced distribution of work among the processors. The domain defining grid is also used to redistribute the elements and points after grid generation, and during the subsequent mesh improvement. Timings show that the proposed approach is scalable and able to produce large grids of high quality in a modest amount of clocktime. With the proposed parallel grid generator, a major impediment to a completely scalable simulation pipeline (grid generation, solvers, post-processing) has been removed, opening the way for truly large-scale computations using unstructured, body-fitted grids.

Grid Generation↗

Scaling the SIESTA Magnetohydrodynamics Equilibrium Code

We report the results of a scaling effort that increases both the speed and resolution of the SIESTA magnetohydrodynamic equilibrium code. SIESTA is capable of computing three-dimensional plasma equilibria with magnetic islands at high spatial resolutions for toroidally confined plasmas. Starting with a small-scale parallel im- plementation, we identified scale-dependent bottlenecks of the code and developed scalable alternatives for each performance-significant functionality, cumulatively increasing the runtime speed and scalability with the number of processors. The net outcome is an improvement in speed by over ten-fold, utilizing a few thousand processors, en- abling SIESTA to simulate high spatial-resolution scenarios in under an hour for the first time.

Seal, Sudip K↗

Quantum Monte Carlo Endstation for Petascale Computing

The major achievements enabled by QMC Endstation grant include * Performance improvement on clusters of x86 multi-core systems, especially on Cray XT systems * New and improved methods for the wavefunction optimizations * New forms of trial wavefunctions * Implementation of the full application on NVIDIA GPUs using CUDA The scaling studies of QMCPACK on large-scale systems show excellent parallel efficiency up to 216K cores on Jaguarpf (Cray XT5). The GPU implementation shows speedups of 10-15x over the CPU implementation on older generation of x86. We have implemented hybrid OpenMP/MPI scheme in QMC to take advantage of multi-core shared memory processors of petascale systems. Our hybrid scheme has several advantages over the standard MPI-only scheme. * Memory optimized: large read-only data to store one-body orbitals and other shared properties to represent the trial wave function and many-body Hamiltonian can be shared among threads, which reduces the memory footprint of a large-scale problem. * Cache optimized: the data associated with an active Walker are in cache during the compute-intensive drift-diffusion process and the operations on an Walker are optimized for cache reuse. Thread-local objects are used to ensure the data affinity to a thread. * Load balanced: Walkers in an ensemble are evenly distributed among threads and MPI tasks. The two-level parallelism reduces the population imbalance among MPI tasks and reduces the number of point-to-point communications of large messages (serialized objects) for the Walker exchange. * Communication optimized: the communication overhead, especially for the collective operations necessary to determine ET and measure the properties of an ensemble, is significantly lowered by using less MPI tasks. The multiple forms of parallelism afforded by QMC algorithms make them ideal candidates for acceleration in the many-core paradigm. We presented the results of our effort to port the QMCPACK simulation code to the NVIDIA CUDA GPU platform. We restructured the CPU algorithms to express additional parallelism, minimize GPU-CPU communication, and efficiently utilize the GPU memory hierarchy. Using mixed precision on GT200 GPUs and MPI for intercommunication and load balancing, we observe typical full-application speedups of approximately 10x to 15x relative to quad-core Xeon CPUs alone, while reproducing the double-precision CPU results within statistical error. We developed an all-electron quantum Monte Carlo (QMC) method for solids that does not rely on pseudopotentials, and used it to construct a primary ultra-high-pressure calibration based on the equation of state of cubic boron nitride. We computed the static contribution to the free energy with the QMC method and obtained the phonon contribution from density functional theory, yielding a high-accuracy calibration up to 900 GPa usable directly in experiment. We computed the anharmonic Raman frequency shift with QMC simulations as a function of pressure and temperature, allowing optical pressure calibration. In contrast to present experimental approaches, small systematic errors in the theoretical EOS do not increase with pressure, and no extrapolation is needed. This all-electron method is applicable to first-row solids, providing a new reference for ab initio calculations of solids and benchmarks for pseudopotential accuracy. We compared experimental and theoretical results on the momentum distribution and the quasiparticle renormalization factor in sodium. From an x-ray Compton-profile measurement of the valence-electron momentum density, we derived its discontinuity at the Fermi wavevector finding an accurate measure of the renormalization factor that we compared with quantum-Monte-Carlo and G0W0 calculations performed both on crystalline sodium and on the homogeneous electron gas. Our calculated results are in good agreement with the experiment. We have been studying the heat of formation for various Kubas complexes of molecular hydrogen on Ti(1,2)ethylene-nH2 using Diffusion Monte Carlo. This work has been started and is ongoing. We are studying systems involving 1 and 2 Ti bonding sites with up to 10 hydrogen molecules in numerous configurations. This work will establish a benchmark that will test the accuracy of density functional calculations and establish the feasibility of our methods for similar systems.

97 MATHEMATICS AND COMPUTING↗

Skel: Generative Software for Producing Skeletal I/O Applications

Massively parallel computations consist of a mixture of computation, communication, and I/O. As part of the co-design for the inevitable progress towards exascale computing, we must apply lessons learned from past work to succeed in this new age of computing. Of the three components listed above, implementing an effective parallel I/O solution has often been overlooked by application scientists and was usually added to large scale simulations only when existing serial techniques had failed. As scientists teams scaled their codes to run on hundreds of processors, it was common to call on an I/O expert to implement a set of more scalable I/O routines. These routines were easily separated from the calculations and communication, and in many cases, an I/O kernel was derived from the application which could be used for testing I/O performance independent of the application. These I/O kernels developed a life of their own used as a broad measure for comparing different I/O techniques. Unfortunately, as years passed and computation and communication changes required changes to the I/O, the separate I/O kernel used for benchmarking remained static no longer providing an accurate indicator of the I/O performance of the simulation making I/O research less relevant for the application scientists. In this paper we describe a new approach to this problem where I/O kernels are replaced with skeletal I/O applications automatically generated from an abstract set of simulation I/O parameters. We realize this abstraction by leveraging the ADIOS middleware's XML I/O specification with additional runtime parameters. Skeletal applications offer all of the benefits of I/O kernels including allowing I/O optimizations to focus on useful I/O patterns. Moreover, since they are automatically generated, it is easy to produce an updated I/O skeleton whenever the simulation's I/O changes. In this paper we analyze the performance of automatically generated I/O skeletal applications for the S3D and GTS codes. We show that these skeletal applications achieve performance comparable to that of the production applications. We wrap up the paper with a discussion of future changes to make the skeletal application better approximate the actual I/O performed in the simulation.

99 GENERAL AND MISCELLANEOUS↗

A Robust Parallel Distributed State Estimation for Large Scale Distribution Systems

The growing need and interest in real-time monitoring of large distribution networks motivated by the rapid population of renewable sources, EVs and etc. demand a computationally efficient state estimation framework. Furthermore, this paper presents an improved computational framework for implementing a robust state estimator using a multi-core processor. The main contribution of the paper is the proposed computational framework along with two partitioning strategies which enable fast and robust state estimation for large scale radial and/or meshed distribution systems. Formulation of the proposed method and its implementation are described in detail. Performance of the estimator is tested by simulations first using a small 84-bus radial distribution system. Then the method’s scalability is demonstrated by simulations on two very large scale distribution networks one configured radially and the other meshed each containing over 12,500 buses.

42 ENGINEERING↗

TriC: Distributed-memory Triangle Counting by Exploiting the Graph Structure

Graph analytics has emerged as an important tool in the analysis of large scale data from diverse application domains such as social networks, cyber security and bioinformatics. Counting the number of triangles in a graph is a fundamental kernel with several applications such as detecting the community structure of a graph or in identifying important vertices in a graph. The ubiquity of massive datasets is driving the need to scale graph analytics on parallel systems. However, numerous challenges exist in efficiently parallelizing graph algorithms, especially on distributed-memory systems. Irregular memory accesses and communication patterns, low computation to communication ratios, and the need for frequent synchronization are some of the leading challenges. In this paper, we present TriC, our distributed-memory implementation of triangle counting in graphs using the Message Passing Interface (MPI), as a submission to the 2020 GraphChallenge competition. Using a set of synthetic and real-world inputs from the challenge, we demonstrate a speedup of up to 90x relative to previous work on 32 processor-cores of a NERSC Cori node. We also provide details from distributed runs with up to8192 processes along with strong scaling results. The observations presented in this work provide an understanding of the system-level bottlenecks at scale that specifically impact sparse-irregular workloads and will therefore benefit other efforts to parallelize graph algorithms.

Halappanavar, Mahantesh↗

Adapting Wave-front Algorithms to Efficiently Utilize Systems with Deep Communication Hierarchies

Large-scale systems increasingly exhibit a differential between intra-chip and inter-chip communication performance especially in hybrid systems using accelerators. Processorcores on the same socket are able to communicate at lower latencies, and with higher bandwidths, than cores on different sockets either within the same node or between nodes. A key challenge is to efficiently use this communication hierarchy and hence optimize performance. We consider here the class of applications that contains wavefront processing. In these applications data can only be processed after their upstream neighbors have been processed. Similar dependencies result between processors in which communication is required to pass boundary data downstream and whose cost is typically impacted by the slowest communication channel in use. In this work we develop a novel hierarchical wave-front approach that reduces the use of slower communications in the hierarchy but at the cost of additional steps in the parallel computation and higher use of on-chip communications. This tradeoff is explored using a performance model. An implementation using the Reverse-acceleration programming model on the petascale Roadrunner system demonstrates a 27% performance improvement at full system-scale on a kernel application. The approach is generally applicable to large-scale multi-core and accelerated systems where a differential in system communication performance exists.

43 PARTICLE ACCELERATORS↗

Comparative Implementation of High Performance Computing for Power System Dynamic Simulations

Dynamic simulation for transient stability assessment is one of the most important, but intensive, computations for power system planning and operation. Present commercial software is mainly designed for sequential computation to run a single simulation, which is very time consuming with a single processer. The application of High Performance Computing (HPC) to dynamic simulations is very promising in accelerating the computing process by parallelizing its kernel algorithms while maintaining the same level of computation accuracy. This paper describes the comparative implementation of four parallel dynamic simulation schemes in two state-of-the-art HPC environments: Message Passing Interface (MPI) and Open Multi-Processing (OpenMP). These implementations serve to match the application with dedicated multi-processor computing hardware and maximize the utilization and benefits of HPC during the development process.

97 MATHEMATICS AND COMPUTING↗

...And Eat it Too: High Read Performance in Write-Optimized HPC I/O Middleware File Formats

As HPC applications run on increasingly high process counts on larger and larger machines, both the frequency of checkpoints needed for fault tolerance and the resolution and size of Data Analysis Dumps are expected to increase proportionally. In order to maintain an acceptable ratio of time spent performing useful computation work to time spent performing I/O, write bandwidth to the underlying storage system must increase proportionally to this increase in the checkpoint and computation size. Unfortunately, popular scientific self-describing file formats such as netCDF and HDF5 are designed with a focus on portability and flexibility. Extra care and careful crafting of the output structure and API calls is required to optimize for write performance using these APIs. To provide sufficient write bandwidth to continue to support the demands of scientific applications, the HPC community has developed a number of I/O middleware layers, that structure output into write-optimized file formats. However, the obvious concern with any write optimized file format would be a corresponding penalty on reads. In the log-structured filesystem, for example, a file generated by random writes could be written efficiently, but reading the file back sequentially later would result in very poor performance. Simulation results require efficient read-back for visualization and analytics, and though most checkpoint files are never used, the efficiency of a restart is very important in the face of inevitable failures. The utility of write speed improving middleware would be greatly diminished if it sacrificed acceptable read performance. In this paper we examine the read performance of two write-optimized middleware layers on large parallel machines and compare it to reading data natively in popular file formats. The two I/O middleware layer examined in this paper are the Adaptable IO System (ADIOS), a library-based approach developed at Oak Ridge National Laboratory to provide a high-level IO API that can be used in place of netCDF or HDF5 to do much more aggressive write-behind and efficient reordering of data locations within the file; and the Parallel Log-structured Filesystem (PLFS), a stackable FUSE filesystem approach developed at Los Alamos National Laboratory that decouples concurrent writes to improve the speed of checkpoints. Since ADIOS is an I/O componentization that affords selection of different I/O methods at or during runtime, through a single API, users can have access to MPI-IO, Posix-IO, HDF5, netCDF, and staging methods. The ADIOS BP file format is a new log-file format that has a superset of the features of both HDF5 and netCDF, but is designed to be portable and flexible while being optimized for writing. PLFS, on the other hand, is mounted as a stackable filesystem on top of an existing parallel filesystem. Reads or writes to the PLFS filesystem are transparently translated into operations on per-process log files stored in the underlying parallel filesystem. Since PLFS performs this translation without application modification, users can write in HDF5, netCDF, or app-specific file formats and PLFS will store the writes in a set of efficiently written log-formatted files, while presenting the user with a logical 'flat' file on reads. Despite their different approaches, the commonality behind both of these middleware systems is that they both write to a log file format. As shown in previous publications, writes are fully optimized in both systems, sometimes resulting in 100x improvements over writing data in popular file formats. But as mentioned above, writing is only one part of the story. In this paper we examine the read performance of our middleware layers on large parallel machines and compare these to reading data either natively or from other popular file formats. We compare the reading performance in two different scenarios: (1) Reading back restarts from the same number of processors as wrote the data and (2) Reading back restart data from a different number of processors. We observe that not only can write-optimized I/O middleware be built to not penalize read speeds, but for important workloads, techniques that improve write performance can, perhaps counterintuitively, improve read speeds over reading from a contiguously organized file format. In the remainder of this paper, we investigate this further through case studies of PLFS and ADIOS on simulation checkpoint restarts.

99 GENERAL AND MISCELLANEOUS↗

Development of parallel DEM for the open source code MFIX

The paper presents the development of a parallel Discrete Element Method (DEM) solver for the open source code, Multiphase Flow with Interphase eXchange (MFIX) based on the domain decomposition method. The performance of the code was evaluated by simulating a bubbling fluidized bed with 2.5 million particles. The DEM solver shows strong scalability up to 256 processors with an efficiency of 81%. Further, to analyze weak scaling, the static height of the fluidized bed was increased to hold 5 and 10 million particles. The results show that global communication cost increases with problem size while the computational cost remains constant. Further, the effects of static bed height on the bubble hydrodynamics and mixing characteristics are analyzed.

42 ENGINEERING↗