Search NASA⌕ Search

SEARCH · Search NASA

Results for “Distributed memory”

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 343 records · Page 19

Asynchronous In Situ Connected-Components Analysis for Complex Fluid flows, In: ISAV '16 Proceedings of the 2nd Workshop on In Situ Infrastructures for Enabling Extreme-scale Analysis and Visualization

The simulation of multiscale physics is an important challenge for scientific computing. For this class of problem, large three-dimensional simulations are performed to advance scientific inquiry. On massively parallel computing systems, the volume of data generated by such approaches can become a productivity bottleneck if the raw data generated from the simulation is analyzed in a post-processing step. To address this, we present a physics-based framework for in situ data reduction that is theoretically grounded in multiscale averaging theory. We show how task parallelism can be exploited to concurrently perform a variety of analysis tasks with data-dependent costs, including the generation of iso-surfaces, morphological analyses, and connected components analysis. All analyses are performed in parallel using distributed memory and use the same domain decomposition as the simulation. A task management framework is constructed to leverage available parallelism within a node for analysis. The capabilities of the framework are to launch asynchronous analysis threads, manage dependencies between different tasks, promote data locality and minimize the impact of data transfers. The framework is applied to analyze GPU-based simulations of two-fluid-phase flow in porous media, generating a set of averaged measures that represents the overall system behavior. We demonstrate how the approach can be applied to perform physically-consistent analysis over fluid sub-regions determined from connected components analysis. Simulations performed on Oak Ridge National Lab's Titan supercomputer are profiled to demonstrate the performance of the associated multi-threaded in situ analysis approach for typical production simulation of two-fluid-phase flow.

Computer Science↗

High performance multivariate visual data exploration for extremely large data

One of the central challenges in modern science is the need to quickly derive knowledge and understanding from large, complex collections of data. We present a new approach that deals with this challenge by combining and extending techniques from high performance visual data analysis and scientific data management. This approach is demonstrated within the context of gaining insight from complex, time-varying datasets produced by a laser wakefield accelerator simulation. Our approach leverages histogram-based parallel coordinates for both visual information display as well as a vehicle for guiding a data mining operation. Data extraction and subsetting are implemented with state-of-the-art index/query technology. This approach, while applied here to accelerator science, is generally applicable to a broad set of science applications, and is implemented in a production-quality visual data analysis infrastructure. We conduct a detailed performance analysis and demonstrate good scalability on a distributed memory Cray XT4 system.

Rübel, O↗

MPI-FAUN: An MPI-Based Framework for Alternating-Updating 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 parallel algorithms to solve the problem for big data sets. The main contribution of this work is a new, high-performance parallel computational framework for a broad class of NMF algorithms that iteratively solves 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). The framework is flexible and able to leverage a variety of NMF and NLS algorithms, including Multiplicative Update, Hierarchical Alternating Least Squares, and Block Principal Pivoting. Our implementation allows us to benchmark and compare different algorithms on massive dense and sparse data matrices of size that spans from few hundreds of millions to billions. We demonstrate the scalability of our algorithm and compare it with baseline implementations, showing significant performance improvements. The code and the datasets used for conducting the experiments are available online.

2D↗

Select Advances in Computational Accelerator Physics

Computational accelerator physics has changed and broadened over the last decade or so. Part of the change is due to the advent of multiple ways of parallel computing. Another part comes from algorithmic developments. The multiple ways of parallel computing include distributed memory parallelism and on-chip parallelism, with the latter coming from architectures (CPU and GPU) having multiple processing elements (cores or streaming multiprocessors) and wide vector (SIMD) instruction units. The basics of these new architectures and their application to computational accelerator physics are briefly reviewed. Algorithmic advances in the select areas of spin tracking, cavity calculations, plasma acceleration, and electron cooling are also reviewed. In some cases the algorithms provide increased fidelity improving the overall accuracy, while in other cases, such as controlled dispersion, the algorithms provide increased fidelity by better modeling the essential physical interaction. Finally, the use of computational frameworks, which provide the basic computational infrastructure, while allowing the capability developer to concentrate on the math and physics, is reviewed in the context of the Vorpal application, which has found use across accelerator physics and many other fields.

Engineering↗

The PetscSF Scalable Communication Layer

PetscSF, the communication component of the Portable, Extensible Toolkit for Scientific Computation (PETSc), is designed to provide PETSc's communication infrastructure suitable for exascale computers that utilize GPUs and other accelerators. PetscSF provides a simple application programming interface (API) for managing common communication patterns in scientific computations by using a star-forest graph representation. PetscSF supports several implementations based on MPI and NVSHMEM, whose selection is based on the characteristics of the application or the target architecture. An efficient and portable model for network and intra-node communication is essential for implementing large-scale applications. The Message Passing Interface, which has been the de facto standard for distributed memory systems, has developed into a large complex API that does not yet provide high performance on the emerging heterogeneous CPU-GPU-based exascale systems. Here, we discuss the design of PetscSF, how it can overcome some difficulties of working directly with MPI on GPUs, and we demonstrate its performance, scalability, and novel features.

97 MATHEMATICS AND COMPUTING↗

Hybrid Parallelism for Volume Rendering on Large-, Multi-, and Many-Core Systems

With the computing industry trending towards multi- and many-core processors, we study how a standard visualization algorithm, ray-casting volume rendering, can benefit from a hybrid parallelism approach. Hybrid parallelism provides the best of both worlds: using distributed-memory parallelism across a large numbers of nodes increases available FLOPs and memory, while exploiting shared-memory parallelism among the cores within each node ensures that each node performs its portion of the larger calculation as efficiently as possible. We demonstrate results from weak and strong scaling studies, at levels of concurrency ranging up to 216,000, and with datasets as large as 12.2 trillion cells. The greatest benefit from hybrid parallelism lies in the communication portion of the algorithm, the dominant cost at higher levels of concurrency. We show that reducing the number of participants with a hybrid approach significantly improves performance.

97 MATHEMATICS AND COMPUTING↗

Multistep inversion workflow for 3D long-offset damped elastic waves in the Fourier domain

We present a new workflow for imaging damped three-dimensional elastic wavefields in the Fourier domain. The workflow employs a multiscale imaging approach, in which offset lengths are laddered, where frequency content and damping of the data are changed cyclically. Thus, the inversion process is launched using short-offset and low-frequency data to recover the long spatial wavelength of the image at a shallow depth. Increasing frequency and offset length leads to the recovery of the fine-scale features of the model at greater depths. For the fixed offset, we employ (in the imaging process) a few discrete frequencies with a set of Laplace damping parameters. The forward problem is solved with a finite-difference frequency-domain method based on a massively parallel iterative solver. The inversion code is based upon the solution of a least squares optimisation problem and is solved using a nonlinear gradient method. It is fully parallelised for distributed memory computational platforms. Our full-waveform inversion workflow is applied to the 3D Marmousi-2 and SEG/EAGE Salt models with long-offset data. The maximum inverted frequencies are 6 Hz for the Marmousi model and 2 Hz for the SEG/EAGE Salt model. The detailed structures are imaged successfully up to the depth approximately equal to one-third of the maximum offset length at a resolution consistent with the inverted frequencies.

3D full waveform inversion↗

Work Stealing and Persistence-based Load Balancers for Iterative Overdecomposed Applications

Applications often involve iterative execution of identical or slowly evolving calculations. Such applications require good initial load balance coupled with efficient periodic rebalancing. In this paper, we consider the design and evaluation of two distinct approaches to addressing this challenge: persistence-based load balancing and work stealing. The work to be performed is overdecomposed into tasks, enabling automatic rebalancing by the middleware. We present a hierarchical persistence-based rebalancing algorithm that performs localized incremental rebalancing. We also present an active-message-based retentive work stealing algorithm optimized for iterative applications on distributed memory machines. These are shown to incur low overheads and achieve over 90% efficiency on 76,800 cores.

persistence-based load balancing↗

Application Characterization at Scale: Lessons learned from developing a distributed Open Community Runtime system for High Performance Computing

Since 2012, the U.S. Department of Energy’s X-Stack program has been developing solutions including runtime systems, programming models, languages, compilers, and tools for the Exascale system software to address crucial performance and power requirements. Fine grain programming models and runtime systems show a great potential to efficiently utilize the underlying hardware. Thus, they are essential to many X-Stack efforts. An abundant amount of small tasks can better utilize the vast parallelism available on current and future machines. Moreover, finer tasks can recover faster and adapt better, due to a decrease in state and control. Nevertheless, current applications have been written to exploit old paradigms (such as Communicating Sequential Processor and Bulk Synchronous Parallel processing). To fully utilize the advantages of these new systems, applications need to be adapted to these new paradigms. As part of the applications’ porting process, in-depth characterization studies, focused on both application characteristics and runtime features, need to take place to fully understand the application performance bottlenecks and how to resolve them. This paper presents a characterization study for a novel high performance runtime system, called the Open Community Runtime, using key HPC kernels as its vehicle. This study has the following contributions: one of the first high performance, fine grain, distributed memory runtime system implementing the OCR standard (version 0.99a); and a characterization study of key HPC kernels in terms of runtime primitives running on both intra and inter node environments. Running on a general purpose cluster, we have found up to 1635x relative speed-up for a parallel tiled Cholesky Kernels on 128 nodes with 16 cores each and a 1864x relative speed-up for a parallel tiled Smith-Waterman kernel on 128 nodes with 30 cores.

Application characterization↗

What does fault tolerant Deep Learning need from MPI?

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

Amatya, Vinay C.↗

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. In this paper, we present the conceptual simulation framework, algorithmic foundations, and runtime interface of CloneX, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a large parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. Finally, the results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the CloneX interface.

97 MATHEMATICS AND COMPUTING↗

Integrated Model, Batch and Domain Parallelism in Training Neural Networks

We propose a new integrated method of exploiting model, batch and domain parallelism for the training of deep neural networks (DNNs) on large distributed-memory computers using minibatch stochastic gradient descent (SGD). Our goal is to find an efficient parallelization strategy for a fixed batch size using $P$ processes. Our method is inspired by the communication-avoiding algorithms in numerical linear algebra. We see $P$ processes as logically divided into a $P_r \times P_c$ grid where the $P_r$ dimension is implicitly responsible for model/domain parallelism and the $P_c$ dimension is implicitly responsible for batch parallelism. In practice, the integrated matrix-based parallel algorithm encapsulates these types of parallelism automatically. We analyze the communication complexity and analytically demonstrate that the lowest communication costs are often achieved neither with pure model nor with pure data parallelism. We also show how the domain parallel approach can help in extending the theoretical scaling limit of the typical batch parallel method.

97 MATHEMATICS AND COMPUTING↗

A communication-avoiding 3D sparse triangular solver

We present a novel distributed memory algorithm to improve the strong scalability of the solution of a sparse triangular system. This operation appears in the solve phase of direct methods for solving general sparse linear systems, Ax = b. Our 3D sparse triangular solver employs several techniques, including a 3D MPI process grid, elimination tree parallelism, and data replication, all of which reduce the per-process communication when combined. We present analytical models to understand the communication cost of our algorithm and show that our 3D sparse triangular solver can reduce the per-process communication volume asymptotically by a factor of O(n1/4) and O(n1/6) for problems arising from the finite element discretizations of 2D "planar" and 3D "non-planar" PDEs, respectively. We implement our algorithm for use in SuperLU_DIST3D, using a hybrid MPI+OpenMP programming model. Our 3D triangular solve algorithm, when run on 12k cores of Cray XC30, outperforms the current state-of-the-art 2D algorithm by 7.2x for planar and 2.7x for the non-planar sparse matrices, respectively.

Sao, Piyush↗

Performance Models for Data Transfers: A Case Study with Molecular Chemistry Kernels

In distributed memory systems, it is paramount to develop strategies to overlap the data transfers between memory nodes with the computations in order to exploit the full potential of these systems. In this paper, we consider the problem of determining the order of data transfers between two memory nodes for a set of independent tasks with the objective of minimizing the makespan. We prove that, with limited memory capacity, the problem of obtaining the optimal order of data transfers is NP-complete. We propose several heuristics to determine this order and discuss the conditions that might be favorable to different heuristics. We analyze our heuristics on traces obtained by running 2 molecular chemistry kernels, namely, Hartree–Fock (HF) and Coupled Cluster Single Double (CCSD) on 10 nodes of an HPC system. Our results show that some of our heuristics achieve significant overlap for moderate memory capacities and resulting in makespans that are very close to the lower bound.

Kumar, Suraj↗

Simulating the Impact of Dynamic Rerouting on Metropolitan-scale Traffic Systems

The rapid introduction of mobile navigation aides that use real-time road network information to suggest alternate routes to drivers is making it more difficult for researchers and government transportation agencies to understand and predict the dynamics of congested transportation systems. Computer simulation is a key capability for these organizations to analyze hypothetical scenarios; however, the complexity of transportation systems makes it challenging for them to simulate very large geographical regions, such as multi-city metropolitan areas. In this article, we describe enhancements to the Mobiliti parallel traffic simulator to model dynamic rerouting behavior with the addition of vehicle controller actors and vehicle-to-controller reroute requests. The simulator is designed to support distributed-memory parallel execution using discrete event simulation and be scalable on high-performance computing platforms. We demonstrate the potential of the simulator by analyzing the impact of varying the population penetration rate of dynamic rerouting on the San Francisco Bay Area road network. Using high-performance parallel computing, we can simulate a day in the San Francisco Bay Area with 19 million vehicle trips with 50 percent dynamic rerouting penetration over a road network with 0.5 million nodes and 1 million links in less than three minutes. We present a sensitivity study on the dynamic rerouting parameters, discuss the simulator’s parallel scalability, and analyze system-level impacts of changing the dynamic rerouting penetration. Furthermore, we examine the varying effects on different functional classes and geographical regions and present a validation of the simulation results compared to real-world data.

97 MATHEMATICS AND COMPUTING↗

Performance Analysis and Optimal Node-aware Communication for Enlarged Conjugate Gradient Methods

Krylov methods are a key way of solving large sparse linear systems of equations but suffer from poor strong scalability on distributed memory machines. Furthermore, this is due to high synchronization costs from large numbers of collective communication calls alongside a low computational workload. Enlarged Krylov methods address this issue by decreasing the total iterations to convergence, an artifact of splitting the initial residual and resulting in operations on block vectors. In this article, we present a performance study of an enlarged Krylov method, Enlarged Conjugate Gradients (ECG), noting the impact of block vectors on parallel performance at scale. Most notably, we observe the increased overhead of point-to-point communication as a result of denser messages in the sparse matrix-block vector multiplication kernel. Additionally, we present models to analyze expected performance of ECG, as well as motivate design decisions. Most importantly, we introduce a new point-to-point communication approach based on node-aware communication techniques that increases efficiency of the method at scale.

97 MATHEMATICS AND COMPUTING↗

Communication Lower Bounds and Optimal Algorithms for Symmetric Matrix Computations

In this article, we focus on the communication costs of three symmetric matrix computations: (i) multiplying a matrix with its transpose, known as a symmetric rank-k update (SYRK) (ii) adding the result of the multiplication of a matrix with the transpose of another matrix and the transpose of that result, known as a symmetric rank-2k update (SYR2K) (iii) performing matrix multiplication with a symmetric input matrix (SYMM). All three computations appear in the Level 3 Basic Linear Algebra Subroutines (BLAS) and have wide use in applications involving symmetric matrices. We establish communication lower bounds for these kernels using sequential and distributed-memory parallel computational models, and we show that our bounds are tight by presenting communication-optimal algorithms for each setting. Our lower bound proofs rely on applying a geometric inequality for symmetric computations and analytically solving constrained nonlinear optimization problems. As a result, the symmetric matrix and its corresponding computations are accessed and performed according to a triangular block partitioning scheme in the optimal algorithms.

Al Daas, Hussam [Rutherford Appleton Laboratory, D↗

Multicore Challenges and Benefits for High Performance Scientific Computing

Until recently, performance gains in processors were achieved largely by improvements in clock speeds and instruction level parallelism. Thus, applications could obtain performance increases with relatively minor changes by upgrading to the latest generation of computing hardware. Currently, however, processor performance improvements are realized by using multicore technology and hardware support for multiple threads within each core, and taking full advantage of this technology to improve the performance of applications requires exposure of extreme levels of software parallelism. We will here discuss the architecture of parallel computers constructed from many multicore chips as well as techniques for managing the complexity of programming such computers, including the hybrid message-passing/multi-threading programming model. We will illustrate these ideas with a hybrid distributed memory matrix multiply and a quantum chemistry algorithm for energy computation using Møller–Plesset perturbation theory.

Nielsen, Ida M. B.↗