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 1,405 records · Page 78

A pervasive parallel framework for visualization: final report for FWP 10-014707

We are on the threshold of a transformative change in the basic architecture of highperformance computing. The use of accelerator processors, characterized by large core counts, shared but asymmetrical memory, and heavy thread loading, is quickly becoming the norm in high performance computing. These accelerators represent significant challenges in updating our existing base of software. An intrinsic problem with this transition is a fundamental programming shift from message passing processes to much more fine thread scheduling with memory sharing. Another problem is the lack of stability in accelerator implementation; processor and compiler technology is currently changing rapidly. This report documents the results of our three-year ASCR project to address these challenges. Our project includes the development of the Dax toolkit, which contains the beginnings of new algorithms for a new generation of computers and the underlying infrastructure to rapidly prototype and build further algorithms as necessary.

97 MATHEMATICS AND COMPUTING↗

A Lanczos eigenvalue method on a parallel computer

Eigenvalue analyses of complex structures is a computationally intensive task which can benefit significantly from new and impending parallel computers. This study reports on a parallel computer implementation of the Lanczos method for free vibration analysis. The approach used here subdivides the major Lanczos calculation tasks into subtasks and introduces parallelism down to the subtask levels such as matrix decomposition and forward/backward substitution. The method was implemented on a commercial parallel computer and results were obtained for a long flexible space structure. While parallel computing efficiency is problem and computer dependent, the efficiency for the Lanczos method was good for a moderate number of processors for the test problem. The greatest reduction in time was realized for the decomposition of the stiffness matrix, a calculation which took 70 percent of the time in the sequential program and which took 25 percent of the time on eight processors. For a sample calculation of the twenty lowest frequencies of a 486 degree of freedom problem, the total sequential computing time was reduced by almost a factor of ten using 16 processors.

Bostic, Susan W.↗

Parallel pseudospectral domain decomposition techniques

The influence of interface boundary conditions on the ability to parallelize pseudospectral multidomain algorithms is investigated. Using the properties of spectral expansions, a novel parallel two domain procedure is generalized to an arbitrary number of domains each of which can be solved on a separate processor. This interface boundary condition considerably simplifies influence matrix techniques.

Gottlieb, David↗

Explicit and implicit solution of the Navier-Stokes equations on a massively parallel computer

The design, implementation, and performance of a two-dimensional time-accurate Navier-Stokes solver for the CM2 supercomputer are described. The program uses a single processor for each grid point. Two different time-stepping methods have so far been implemented: an explicit third-order Runge-Kutta method and an implicit approximation-factorization method. The CM2 results are checked against those of a mature well-vectorized Cray 2 program, both for correctness and performance. The code is found to be correct, and the performance in some cases is up to several times that of the Cray 2.

Levit, Creon↗

Parallel pseudospectral domain decomposition techniques

The influence of interface boundary conditions on the ability to parallelize pseudospectral multidomain algorithms is investigated. Using the properties of spectral expansions, a novel parallel two domain procedure is generalized to an arbitrary number of domains each of which can be solved on a separate processor. This interface boundary condition considerably simplifies influence matrix techniques.

Gottlieb, David↗

Performance of a plasma fluid code on the Intel parallel computers

One approach to improving the real-time efficiency of plasma turbulence calculations is to use a parallel algorithm. A parallel algorithm for plasma turbulence calculations was tested on the Intel iPSC/860 hypercube and the Touchtone Delta machine. Using the 128 processors of the Intel iPSC/860 hypercube, a factor of 5 improvement over a single-processor CRAY-2 is obtained. For the Touchtone Delta machine, the corresponding improvement factor is 16. For plasma edge turbulence calculations, an extrapolation of the present results to the Intel (sigma) machine gives an improvement factor close to 64 over the single-processor CRAY-2.

Lynch, V. E.↗

NbN A/D Conversion of IR Focal Plane Sensor Signal at 10 K

We are implementing a 12 bit SFQ counting ADC with parallel-to-serial readout using our established 10 K NbN capability. This circuit provides a key element of the analog signal processor (ASP) used in large infrared focal plane arrays. The circuit processes the signal data stream from a Si:As BIB detector array. A 10 mega samples per second (MSPS) pixel data stream flows from the chip at a 120 megabit bit rate in a format that is compatible with other superconductive time dependent processor (TDP) circuits being developed. We will discuss our planned ASP demonstration, the circuit design, and test results.

conversion↗

DMA shared byte counters in a parallel computer

A parallel computer system is constructed as a network of interconnected compute nodes. Each of the compute nodes includes at least one processor, a memory and a DMA engine. The DMA engine includes a processor interface for interfacing with the at least one processor, DMA logic, a memory interface for interfacing with the memory, a DMA network interface for interfacing with the network, injection and reception byte counters, injection and reception FIFO metadata, and status registers and control registers. The injection FIFOs maintain memory locations of the injection FIFO metadata memory locations including its current head and tail, and the reception FIFOs maintain the reception FIFO metadata memory locations including its current head and tail. The injection byte counters and reception byte counters may be shared between messages.

97 MATHEMATICS AND COMPUTING↗

RGG: Reactor geometry (and mesh) generator

The reactor geometry (and mesh) generator RGG takes advantage of information about repeated structures in both assembly and core lattices to simplify the creation of geometry and mesh. It is released as open source software as a part of the MeshKit mesh generation library. The methodology operates in three stages. First, assembly geometry models of various types are generated by a tool called AssyGen. Next, the assembly model or models are meshed by using MeshKit tools or the CUBIT mesh generation tool-kit, optionally based on a journal file output by AssyGen. After one or more assembly model meshes have been constructed, a tool called CoreGen uses a copy/move/merge process to arrange the model meshes into a core model. In this paper, we present the current state of tools and new features in RGG. We also discuss the parallel-enabled CoreGen, which in several cases achieves super-linear speedups since the problems fit in available RAM at higher processor counts. Several RGG applications - 1/6 VHTR model, 1/4 PWR reactor core, and a full-core model for Monju - are reported. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Implementation and analysis of a Navier-Stokes algorithm on parallel computers

The results of the implementation of a Navier-Stokes algorithm on three parallel/vector computers are presented. The object of this research is to determine how well, or poorly, a single numerical algorithm would map onto three different architectures. The algorithm is a compact difference scheme for the solution of the incompressible, two-dimensional, time-dependent Navier-Stokes equations. The computers were chosen so as to encompass a variety of architectures. They are the following: the MPP, an SIMD machine with 16K bit serial processors; Flex/32, an MIMD machine with 20 processors; and Cray/2. The implementation of the algorithm is discussed in relation to these architectures and measures of the performance on each machine are given. The basic comparison is among SIMD instruction parallelism on the MPP, MIMD process parallelism on the Flex/32, and vectorization of a serial code on the Cray/2. Simple performance models are used to describe the performance. These models highlight the bottlenecks and limiting factors for this algorithm on these architectures. Finally, conclusions are presented.

Fatoohi, Raad A.↗

Improving Uintah's Scalability Through the Use of Portable Kokkos-Based Data Parallel Tasks

The University of Utah’s Carbon Capture Multidisciplinary Simulation Center (CCMSC) is using the Uintah Computational Framework to predict performance of a 1000 MWe ultra-supercritical clean coal boiler. The center aims to utilize the Intel Xeon Phi-based DOE systems, Theta and Aurora, through the Aurora Early Science Program by using the Kokkos C++ library to enable node-level performance portability. This paper describes infrastructure advancements and portability improvements made possible by the integration of Kokkos within Uintah. This integration marks a step towards consolidating Uintah’s MPI+PThreads and MPI+CUDA hybrid parallelism approaches into a single MPI+Kokkos approach. Scalability results are presented that compare serial and data parallel task execution models for a challenging radiative heat transfer calculation, central to the center’s predictive boiler simulations. Here, these results demonstrate both good strong-scaling characteristics to 256 Knights Landing (KNL) processors on the NSF Stampede system, and show the KNL-based calculation to compete with prior GPU-based results for the same calculation.

97 MATHEMATICS AND COMPUTING↗

(abstract) A High Throughput 3-D Inner Product Processor

A particularily challenging image processing application is the real time scene acquisition and object discrimination. It requires spatio-temporal recognition of point and resolved objects at high speeds with parallel processing algorithms. Neural network paradigms provide fine grain parallism and, when implemented in hardware, offer orders of magnitude speed up. However, neural networks implemented on a VLSI chip are planer architectures capable of efficient processing of linear vector signals rather than 2-D images. Therefore, for processing of images, a 3-D stack of neural-net ICs receiving planar inputs and consuming minimal power are required. Details of the circuits with chip architectures will be described with need to develop ultralow-power electronics. Further, use of the architecture in a system for high-speed processing will be illustrated.

imaging parallel processing algorithms linear vect↗

Gigaflop performance on a CRAY-2: Multitasking a computational fluid dynamics application

The methodology is described for converting a large, long-running applications code that executed on a single processor of a CRAY-2 supercomputer to a version that executed efficiently on multiple processors. Although the conversion of every application is different, a discussion of the types of modification used to achieve gigaflop performance is included to assist others in the parallelization of applications for CRAY computers, especially those that were developed for other computers. An existing application, from the discipline of computational fluid dynamics, that had utilized over 2000 hrs of CPU time on CRAY-2 during the previous year was chosen as a test case to study the effectiveness of multitasking on a CRAY-2. The nature of dominant calculations within the application indicated that a sustained computational rate of 1 billion floating-point operations per second, or 1 gigaflop, might be achieved. The code was first analyzed and modified for optimal performance on a single processor in a batch environment. After optimal performance on a single CPU was achieved, the code was modified to use multiple processors in a dedicated environment. The results of these two efforts were merged into a single code that had a sustained computational rate of over 1 gigaflop on a CRAY-2. Timings and analysis of performance are given for both single- and multiple-processor runs.

Tennille, Geoffrey M.↗

Parallel-vector algorithms for particle simulations on shared-memory multiprocessors

Over the last few decades, the computational demands of massive particle-based simulations for both scientific and industrial purposes have been continuously increasing. Hence, considerable efforts are being made to develop parallel computing techniques on various platforms. In such simulations, particles freely move within a given space, and so on a distributed-memory system, load balancing, i.e., assigning an equal number of particles to each processor, is not guaranteed. However, shared-memory systems achieve better load balancing for particle models, but suffer from the intrinsic drawback of memory access competition, particularly during (1) paring of contact candidates from among neighboring particles and (2) force summation for each particle. Here, novel algorithms are proposed to overcome these two problems. For the first problem, the key is a pre-conditioning process during which particle labels are sorted by a cell label in the domain to which the particles belong. Then, a list of contact candidates is constructed by pairing the sorted particle labels. For the latter problem, a table comprising the list indexes of the contact candidate pairs is created and used to sum the contact forces acting on each particle for all contacts according to Newton's third law. With just these methods, memory access competition is avoided without additional redundant procedures. The parallel efficiency and compatibility of these two algorithms were evaluated in discrete element method (DEM) simulations on four types of shared-memory parallel computers: a multicore multiprocessor computer, scalar supercomputer, vector supercomputer, and graphics processing unit. The computational efficiency of a DEM code was found to be drastically improved with our algorithms on all but the scalar supercomputer. Thus, the developed parallel algorithms are useful on shared-memory parallel computers with sufficient memory bandwidth.

97 MATHEMATICS AND COMPUTING↗

A Case Study of LLVM-Based Analysis for Optimizing SIMD Code Generation

This paper presents a methodology for using LLVM-based tools to tune the DCA++ (dynamical cluster approximation) application that targets the new ARM A64FX processor. The goal is to describe the changes required for the new architecture and generate efficient single instruction/multiple data (SIMD) instructions that target the new Scalable Vector Extension instruction set. During manual tuning, the authors used the LLVM tools to improve code parallelization by using OpenMP SIMD, refactored the code and applied transformation that enabled SIMD optimizations, and ensured that the correct libraries were used to achieve optimal performance. By applying these code changes, code speed was increased by 1.98× and 78 GFlops were achieved on the A64FX processor. The authors aim to automatize parts of the efforts in the OpenMP Advisor tool, which is built on top of existing and newly introduced LLVM tooling.

Huber, Joseph↗

CosmoFlow: Using Deep Learning to Learn the Universe at Scale

Deep learning is a promising tool to determine the physical model that describes our universe. To handle the considerable computational cost of this problem, we present CosmoFlow: a highly scalable deep learning application built on top of the TensorFlow framework. CosmoFlow uses efficient implementations of 3D convolution and pooling primitives, together with improvements in threading for many element-wise operations, to improve training performance on Intel ® Xeon Phi™ processors. We also utilize the Cray PE Machine Learning Plugin for efficient scaling to multiple nodes. We demonstrate fully synchronous data-parallel training on 8192 nodes of Cori with 77% parallel efficiency, achieving 3.5 Pflop/s sustained performance. To our knowledge, this is the first large-scale science application of the TensorFlow framework at supercomputer scale with fully-synchronous training. Here, these enhancements enable us to process large 3D dark matter distribution and predict the cosmological parameters Ω M , σ 8 and n s with unprecedented accuracy.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Improved parallel performance of the CICE model in CESM1

The Los Alamos sea ice model, CICE, is a sophisticated finite difference grid point model. It has been a part of the National Science Foundation and Department of Energy community climate models (Community Climate System Model (CCSM) and Community Earth System Model (CESM)) for over a decade. It includes various physical and dynamical processes and is parallelized to run on large-scale computer systems. The CICE model was assessed in the CESM at different resolutions and target processor counts to better understand and optimize the performance. Several new decompositions and a new feature to reduce the halo cost were added to the model. The new decompositions better leverage land block elimination and take advantage of scaling opportunities in different computational kernels. As a result of these new features, the CICE model performance has been improved by up to 45% and has more flexibility to be run efficiently at arbitrary Message Passing Interface (MPI) task counts

54 ENVIRONMENTAL SCIENCES↗

Efficiently modeling neural networks on massively parallel computers

Neural networks are a very useful tool for analyzing and modeling complex real world systems. Applying neural network simulations to real world problems generally involves large amounts of data and massive amounts of computation. To efficiently handle the computational requirements of large problems, we have implemented at Los Alamos a highly efficient neural network compiler for serial computers, vector computers, vector parallel computers, and fine grain SIMD computers such as the CM-2 connection machine. This paper describes the mapping used by the compiler to implement feed-forward backpropagation neural networks for a SIMD (Single Instruction Multiple Data) architecture parallel computer. Thinking Machines Corporation has benchmarked our code at 1.3 billion interconnects per second (approximately 3 gigaflops) on a 64,000 processor CM-2 connection machine (Singer 1990). This mapping is applicable to other SIMD computers and can be implemented on MIMD computers such as the CM-5 connection machine. Our mapping has virtually no communications overhead with the exception of the communications required for a global summation across the processors (which has a sub-linear runtime growth on the order of O(log(number of processors)). We can efficiently model very large neural networks which have many neurons and interconnects and our mapping can extend to arbitrarily large networks (within memory limitations) by merging the memory space of separate processors with fast adjacent processor interprocessor communications. This paper will consider the simulation of only feed forward neural network although this method is extendable to recurrent networks.

Farber, Robert M.↗