Search NASA⌕ Search

SEARCH · Search NASA

Results for “PARALLEL PROCESSING”

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 1,225 records · Page 68

Accelerating Numerical Calculation on the Cray XMT

The Cray XMT provides hardware support for parallel algorithms that would be communication- or memory-bound on other machines. Unfortunately, even if an algorithm meets these criteria, performance suffers if the algorithm is too numerically intensive. We present a lookup-based approach that achieves a significant performance advantage over explicit calculation. We describe an approach to balancing memory bandwidth against on-chip floating point capabilities, leading to further speedup. Finally, we provide table lookup algorithms for a number of common functions.

97 MATHEMATICS AND COMPUTING↗

: A Scalable and Transparent System for Simulating MPI Programs

is a scalable, transparent system for experimenting with the execution of parallel programs on simulated computing platforms. The level of simulated detail can be varied for application behavior as well as for machine characteristics. Unique features of are repeatability of execution, scalability to millions of simulated (virtual) MPI ranks, scalability to hundreds of thousands of host (real) MPI ranks, portability of the system to a variety of host supercomputing platforms, and the ability to experiment with scientific applications whose source-code is available. The set of source-code interfaces supported by is being expanded to support a wider set of applications, and MPI-based scientific computing benchmarks are being ported. In proof-of-concept experiments, has been successfully exercised to spawn and sustain very large-scale executions of an MPI test program given in source code form. Low slowdowns are observed, due to its use of purely discrete event style of execution, and due to the scalability and efficiency of the underlying parallel discrete event simulation engine, sik. In the largest runs, has been executed on up to 216,000 cores of a Cray XT5 supercomputer, successfully simulating over 27 million virtual MPI ranks, each virtual rank containing its own thread context, and all ranks fully synchronized by virtual time.

97 MATHEMATICS AND COMPUTING↗

Effects of floating-point non-associativity on numerical computations on massively multithreaded systems

Floating-point addition and multiplication are not necessarily associative. When performing those operations over large numbers of operands with different magnitudes, the order in which individual operations are performed can affect the final result. On massively multithreaded systems, when performing parallel reductions, the non-deterministic nature of numerical operation interleaving can lead to non-deterministic numerical results. We have investigated the effect of this problem on the convergence of a conjugate gradient calculation used as part of a power grid analysis application.

97 MATHEMATICS AND COMPUTING↗

Parallel Adaptive Simulation of Weak and Strong Transverse-Wave Structures in H2-O2 Detonations

Two- and three-dimensional simulation results are presented that investigate at great detail the temporal evolution of Mach reflection sub-structure patterns intrinsic to gaseous detonation waves. High local resolution is achieved by utilizing a distributed memory parallel shock-capturing finite volume code that employs block-structured dynamic mesh adaptation. The computational approach, the implemented parallelization strategy, and the software design are discussed.

08 HYDROGEN↗

Three-Dimensional Parallel Adaptive Mesh Refinement Simulations of Shock-Driven Turbulent Mixing in Plane and Converging Geometries

This paper presents the use of a dynamically adaptive mesh refinement strategy for the simulations of shock-driven turbulent mixing. Large-eddy simulations are necessary due the high Reynolds number turbulent regime. In this approach, the large scales are simulated directly and small scales at which the viscous dissipation occurs are modeled. A low-numerical centered finite-difference scheme is used in turbulent flow regions while a shock-capturing method is employed to capture shocks. Three-dimensional parallel simulations of the Richtmyer-Meshkov instability performed in plane and converging geometries are described.

42 ENGINEERING↗

Enabling a Highly-Scalable Global Address Space Model for Petascale Computing

Over the past decade, the trajectory to the petascale has been built on increased complexity and scale of the underlying parallel architectures. Meanwhile, software de- velopers have struggled to provide tools that maintain the productivity of computational science teams using these new systems. In this regard, Global Address Space (GAS) programming models provide a straightforward and easy to use addressing model, which can lead to improved produc- tivity. However, the scalability of GAS depends directly on the design and implementation of the runtime system on the target petascale distributed-memory architecture. In this paper, we describe the design, implementation, and optimization of the Aggregate Remote Memory Copy Interface (ARMCI) runtime library on the Cray XT5 2.3 PetaFLOPs computer at Oak Ridge National Laboratory. We optimized our implementation with the flow intimation technique that we have introduced in this paper. Our optimized ARMCI implementation improves scalability of both the Global Arrays (GA) programming model and a real-world chemistry application NWChem from small jobs up through 180,000 cores.

97 MATHEMATICS AND COMPUTING↗

Reducing Application Runtime Variability on Jaguar XT5

Operating system (OS) noise is defined as interference generated by the OS that prevents a compute core from performing ``useful'' work. Compute node kernel daemons, network interfaces, and other OS related services are major sources of such interference. This interference on individual compute cores can vary in duration and frequency, and can cause de-synchronization (jitter) in collective communication tasks and thus results in variable (degraded) overall parallel application performance. This behavior is more observable in large-scale applications using certain types of collective communication primitives, such as MPI\_Allreduce. This paper presents our effort towards reducing the overall effect of OS noise on our large-scale parallel applications. Our tests were performed on the quad-core Jaguar, the Cray XT5 at the Oak Ridge National Laboratory Leadership Computing Facility (OLCF). At the time of these tests, Jaguar was a 1.4 PFLOPS supercomputer with 149,504 compute cores and 8 cores per node. We aggregated OS noise sources onto a single core for each node. The scientific application was then run on six of the remaining cores in each node. Our results show that we were able to improve the MPI_Allreduce performance by two orders of magnitude. We demonstrated up to a 30% boost in the performance of the Parallel Ocean Program (POP) using this technique.

97 MATHEMATICS AND COMPUTING↗

Initial Characterization of Parallel NFS Implementations

Parallel NFS (pNFS) is touted as an emergent standard protocol for parallel I/O access in various storage environments. Several pNFS prototypes have been implemented for initial validation and protocol examination. Previous efforts have focused on realizing the pNFS protocol to expose the best bandwidth potential from underlying file and storage systems. In this presentation, we provide an initial characterization of two pNFS prototype implementations, lpNFS (a Lustre-based parallel NFS implementation) and spNFS (another reference implementation from Network Appliance, Inc.). We show that both lpNFS and spNFS can faithfully achieve the primary goal of pNFS, i.e., aggregating I/O bandwidth from many storage servers. However, they both face the challenge of scalable metadata management. Particularly, the throughput of sp-NFS metadata operations degrades significanlty with an increasing number of data servers. Even for the better-performing lpNFS, we discuss its architecture and propose a direct I/O request flow protocol to improve its performance.

97 MATHEMATICS AND COMPUTING↗

CIFTS : A coordinated infrastructure for fault-tolerant systems.

In the next few years SciDAC applications will utilize petascale systems with tens to hundreds of thousands of processors, hundreds of I/O nodes, and thousands of disks. This leap of two orders of magnitude in scale from today's typical systems is causing a critical gap in fault management of these systems. The fault management issues for these emerging systems are well beyond the scope of today's common infrastructure and practice. Currently, systems software components for large-scale machines remain largely independent in their fault awareness and notification strategies. Faults can arise not just from the hardware but also from the OS, middleware, libraries, and application levels. Petascale applications that are evolving to utilize these platforms face many new challenges. With the CIFTS initiative, we aim to provide a coordinated infrastructure that will enable Fault Tolerant Systems to adapt to faults occuring in the operating environment in a holistic manner. Our approach will be to design a reference implementation of a fault awareness and notification backplane to provide common uniform event handling and notification mechanisms for fault-aware libraries and middleware; create an interface specification that allows libraries, run- time systems, and applications to connect to and use the fault-tolerance backplane; and extend key libraries and applications to validate the interface choices and to form the critical mass necessary for adoption in the community.

99 GENERAL AND MISCELLANEOUS↗

Analyzing the trade-off between multiple memory controllers and memory channels on multi-core processor performance

Increasing the core-count on current and future processors is posing critical challenges to the memory subsystem to efficiently handle concurrent memory requests. The current trend to cope with this challenge is to increase the number of memory channels available to the processor's memory controller. In this paper we investigate the effectiveness of this approach on the performance of parallel scientific applications. Specifically, we explore the trade-off between employing multiple memory channels per memory controller and the use of multiple memory controllers. Experiments conducted on two current state-of-the-art multicore processors, a 6-core AMD Istanbul and a 4-core Intel Nehalem-EP, for a wide range of production applications shows that there is a diminishing return when increasing the number of memory channels per memory controller. In addition, we show that this performance degradation can be efficiently addressed by increasing the ratio of memory controllers to channels while keeping the number of memory channels constant. Significant performance improvements can be achieved in this scheme, up to 28%, in the case of using two memory controllers with each with one channel compared with one controller with two memory channels.

97 MATHEMATICS AND COMPUTING↗

Denovo--A New Three-Dimensional Parallel Discrete Ordinates Code in SCALE

Denovo is a new, three-dimensional, discrete ordinates (SN) transport code that uses state-of-the-art solution methods to obtain accurate solutions to the Boltzmann transport equation. Denovo uses the Koch-Baker-Alcouffe parallel sweep algorithm to obtain high parallel efficiency on O(100) processors on XYZ orthogonal meshes. As opposed to traditional SN codes that use source iteration, Denovo uses nonstationary Krylov methods to solve the within-group equations. Krylov methods are far more efficient than stationary schemes. Additionally, classic acceleration schemes (diffusion synthetic acceleration) do not suffer stability problems when used as a preconditioner to a Krylov solver. Denovo's generic programming framework allows multiple spatial discretization schemes and solution methodologies. Denovo currently provides diamond-difference, theta-weighted diamond-difference, linear-discontinuous finite element, trilinear-discontinuous finite element, and step characteristics spatial differencing schemes. Also, users have the option of running traditional source iteration instead of Krylov iteration. Multigroup upscatter problems can be solved using Gauss-Seidel iteration with transport, two-grid acceleration. A parallel first-collision source is also available. Denovo solutions to the Kobayashi benchmarks are in excellent agreement with published results. Parallel performance shows excellent weak scaling up to 20000 cores and good scaling up to 40000 cores.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Thermodynamics of Magnetic Systems from First Principles: WL-LSMS

Density Functional calculations have proven to be a powerful tool to study the ground state of many materials. For finite temperatures the situation is less ideal and one is often forced to rely on models with parameters either fitted to zero temperature first principles calculations or experimental results. This approach is especially unsatisfacory in inhomogeneous systems, nano particles, or other systems where the model parameters could vary significantly from one site to another. Here we describe a possible solution to this problem by combining classical Monte Carlo calculations the Wang-Landau method in this case with a firs principles electronic structure calculation, specifically our locally selfconsistent multiple scallering code (LSMS). The combined code shows superb scaling behavior on massively parallel computers. The code sustained 1.836 Petaflop/s on 223232 cores of the Cray XT5 jaguar system at Oak Ridge.

99 GENERAL AND MISCELLANEOUS↗

Parallelism in System Tools

The Cray XT, when employed in conjunction with the Lustre filesystem, has provided the ability to generate huge amounts of data in the form of many files. Typically, this is accommodated by satisfying the requests of large numbers of Lustre clients in parallel. In contrast, a single service node (Lustre client) cannot adequately service such datasets. This means that the use of traditional UNIX tools like cp, tar, et alli (with have no parallel capability) can result in substantial impact to user productivity. For example, to copy a 10 TB dataset from the service node using cp would take about 24 hours, under more or less ideal conditions. During production operation, this could easily extend to 36 hours. In this paper, we introduce the Lustre User Toolkit for Cray XT, developed at the Oak Ridge Leadership Computing Facility (OLCF). We will show that Linux commands, implementing highly parallel I/O algorithms, provide orders of magnitude greater performance, greatly reducing impact to productivity.

97 MATHEMATICS AND COMPUTING↗