检索结果-内蒙古大学图书馆

25th ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC)

作者： Imamura, Toshiyuki Yamada, Susumu Machida, Masahiko Univ Electrocommun Grad Sch Informat & Engn Chofu Tokyo 182 Japan RIKEN Adv Inst Computat Sci Kobe Hyogo Japan Japan Atom Energy Agcy Kashiwa Chiba Japan

ISBN: (纸本)9780769549569;9781467362184

This study covers the design and implementation of a DD (double-double) extended parallel eigenvalue solver, namely QPEigenK. We extended most of underlying numerical software layers from BLAS, LAPACK, and ScaLAPACK as well as MPI. Preliminary results show that QPEigenK performs on several platforms, and shows good accuracy and parallel efficiency. We can conclude that the DD format is reasonable data format instead of real (16) format from the viewpoint of programming and performance.

关键词： double-double data format distributed memory parallelism parallel eigenvalue solver

来源：评论

学校读者我要写书评

暂无评论

Layer potential evaluations on distributed memory machines

Layer potential evaluations on distributed memory machines

引用

作者： Gao, Hao University of Illinois – Urbana-Champaign

学位级别：硕士

One of the main challenges of using integral equation methods (IEM) for solving partial differential equations is evaluating layer potentials with singular kernels. Quadrature by Expansion (QBX) is a quadrature method to evaluate such layer potentials accurately for targets near or on the source boundary, by forming expansions in the high-accuracy region away from the boundary, and evaluating the targets using the expansions. Recently, a new algorithm, called 'GIGAQBX', has combined QBX with the Fast Multipole Method to achieve linear complexity in terms of the number of degrees of freedom. Despite this advancement, QBX is still computationally expensive. To enable IEM on large-scale problems, this thesis investigates evaluating layer potentials on distributed-memory machines. The distributed algorithm introduced in this thesis is based on GIGAQBX and shows GIGAQBX contains plenty of parallelism. We evaluate our algorithm on the Comet supercomputer at the San Diego Supercomputer Center and show that it exhibits good strong scaling up to 1536 cores.

关键词： Layer Potentials distributed memory parallelism Integral Equations Singular Integrals Fast Multipole Method

来源：评论

学校读者我要写书评

暂无评论

Parallelization of a distributed ecohydrological model

引用

ENVIRONMENTAL MODELLING & SOFTWARE 2018年 101卷 51-63页

作者： Liu, Ning Shaikh, Mohsin Ahmed Kala, Jatin Harper, Richard J. Dell, Bernard Liu, Shirong Sun, Ge Murdoch Univ Sch Vet & Life Sci South St Murdoch WA 6150 Australia Chinese Acad Forestry Beijing 10091 Peoples R China Pawsey Supercomp Ctr Perth WA Australia US Forest Serv USDA Southern Res Stn Eastern Forest Environm Threat Assessment Ctr Raleigh NC 27606 USA

WaSSI-C is an ecohydrological model which couples water and carbon cycles with water use efficiency (WUE) derived from global eddy flux observations. However, a significant limitation of the WaSSI-C model is that it only runs serially. High resolution simulations at a large scale are therefore computationally expensive and cause a run-time memory burden. Using distributed (MPI) and shared (OpenMP) memory parallelism techniques, we revised the original model as dWaSSI-C. We showed that using MPI was effective in reducing the computational run-time and memory use. Two experiments were carried out to simulate water and carbon fluxes over the Australian continent to test the sensitivity of the parallelized model to input data-sets of different spatial resolutions, as well as to WUE parameters for different vegetation types. These simulations were completed within minutes using dWaSSI-C, whereas they would not have been possible with the serial version. The dWaSSI-C model was able to simulate the seasonal dynamics of gross ecosystem productivity (GEP) reasonably well when compared to observations at four eddy flux sites. Sensitivity analysis showed that simulated GEP was more sensitive to WUE during the summer compared to winter in Australia, and woody savannas and grasslands showed higher sensitivity than evergreen broadleaf forests and shrublands. Although our results are model-specific, the parallelization approach can be adopted in other similar ecosystem models for large scale applications. (C) 2017 Elsevier Ltd. All rights reserved.

关键词： High performance computing Ecohydrological modeling distributed memory parallelism Shared memory parallelism Water and carbon fluxes

来源：评论

学校读者我要写书评

暂无评论

Massively parallel implementation and approaches to simulate quantum dynamics using Krylov subspace techniques

引用

COMPUTER PHYSICS COMMUNICATIONS 2019年 235卷 477-488页

作者： Brenes, Marlon Varma, Vipin Kerala Scardicchio, Antonello Girotto, Ivan Abdus Salam Int Ctr Theoret Phys Str Costiera 11 I-34151 Trieste Italy Trinity Coll Dublin Sch Phys Dublin 2 Ireland CUNY Grad Ctr Initiat Theoret Sci New York NY 10016 USA CUNY Coll Staten Isl Dept Engn Sci & Phys Staten Isl NY 10314 USA Univ Pittsburgh Dept Phys & Astron Pittsburgh PA 15260 USA INFN Sez Trieste Via Valerio 2 I-34127 Trieste Italy Univ Modena & Reggio Emilia I-41121 Modena Italy Eindhoven Univ Technol NL-5600 MB Eindhoven Netherlands

We have developed an application and implemented parallel algorithms in order to provide a computational framework suitable for massively parallel supercomputers to study the unitary dynamics of quantum systems. We use renowned parallel libraries such as PETSc/SLEPc combined with high-performance computing approaches in order to overcome the large memory requirements to be able to study systems whose Hilbert space dimension comprises over 9 billion independent quantum states. Moreover, we provide descriptions of the parallel approach used for the three most important stages of the simulation: handling the Hilbert subspace basis, constructing a matrix representation for a generic Hamiltonian operator and the time evolution of the system by means of the Krylov subspace methods. We employ our setup to study the evolution of quasidisordered and clean many-body systems, focussing on the return probability and related dynamical exponents: the large system sizes accessible provide novel insights into their thermalization properties. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Strongly interacting systems Krylov subspace methods Unitary quantum dynamics distributed memory parallelism

来源：评论

学校读者我要写书评

暂无评论

Analyzing asynchronous pipeline schedules

引用

INTERNATIONAL JOURNAL OF PARALLEL PROGRAMMING 1998年第1期26卷 5-42页

作者： Donaldson, V Ferrante, J Univ Calif San Diego Dept Comp Sci & Engn La Jolla CA 92093 USA

Asynchronous pipelining is a form of parallelism that is useful in both distributed and shared memory systems. We show that asynchronous pipeline schedules are a generalization of both noniterative DAG (directed acyclic graph) schedules as well as simpler pipeline schedules, unifying these two types of scheduling. We generalize previous work on determining if a pipeline schedule will deadlock, and generalize Reiter's well-known formula((1)) for determining the iteration interval of a deadlock-free schedule, which is the primary measure of the execution time of a schedule. Our generalizations account for nonzero communication times (easy) and the assignment of multiple tasks to processors (nontrivial). A key component of our generalized approach to pipeline schedule analysis is the use of pipeline scheduling edges with potentially negative data dependence distances. We also discuss implementation of an asynchronous pipeline schedule at runtime;show how to efficiently simulate pipeline execution on a sequential processor;define and derive bounds on the startup time of a schedule, which is a secondary schedule performance measure;and describe a nsw algorithm for evaluating the iteration interval formula.

关键词： pipelining loop optimization task scheduling distributed memory parallelism

来源：评论

学校读者我要写书评

暂无评论

Leveraging HPC accelerator architectures with modern techniques - hydrologic modeling on GPUs with ParFlow

引用

COMPUTATIONAL GEOSCIENCES 2021年第5期25卷 1579-1590页

作者： Hokkanen, Jaro Kollet, Stefan Kraus, Jiri Herten, Andreas Hrywniak, Markus Pleiter, Dirk Forschungszentrum Julich Agrosphere IBG 3 Julich Germany NVIDIA GmbH Wurselen Germany NVIDIA GmbH Julich Supercomputing Ctr Julich Germany

Rapidly changing heterogeneous supercomputer architectures pose a great challenge to many scientific communities trying to leverage the latest technology in high-performance computing. Many existing projects with a long development history have resultedin a large amount of code that is not directly compatible with the latest accelerator architectures. Furthermore, due to limited resources of scientific institutions, developing and maintaining architecture-specific ports is generally unsustainable. In order to adapt to modern accelerator architectures, many projects rely on directive-based programming models or build the codebase tightly around a third-party domain-specific language or library. This introduces external dependencies out of control of theproject. The presented paper tackles the issue by proposing a lightweight application-side adaptor layer for compute kernels and memory management resulting in a versatile and inexpensive adaptation of new accelerator architectures with little drawbacks.A widely used hydrologic model demonstrates that such an approach pursued more than 20 years ago is still paying off with modern accelerator architectures as demonstrated by a very significant performance gain from NVIDIA A100 GPUs, high developer productivity, and minimally invasive implementation;all while the codebase is kept well maintainable in the long-term.

关键词： High-performance computing (HPC) GPU computing distributed memory parallelism Accelerator architecture Domain-specific language (DSL) 68– 04 68 N19

来源：评论

学校读者我要写书评

暂无评论

Scalable Adaptive PDE Solvers in Arbitrary Domains 21

Scalable Adaptive PDE Solvers in Arbitrary Domains

引用

International Conference for High Performance Computing, Networking, Storage and Analysis (SC21)

作者： Saurabh, Kumar Ishii, Masado Fernando, Milinda Gao, Boshun Tan, Kendrick Hsu, Ming-Chen Krishnamurthy, Adarsh Sundar, Hari Ganapathysubramanian, Baskar Iowa State Univ Ames IA 50011 USA Univ Utah Salt Lake City UT USA

ISBN: (纸本)9781450384421

Efficiently and accurately simulating partial differential equations (PDEs) in and around arbitrarily defined geometries, especially with high levels of adaptivity, has significant implications for different application domains. A key bottleneck in the above process is the fast construction of :I 'good' adaptively-relined mesh. In this work, we present an efficient novel oclree-based adaptive discrelizalion approach capable of carving out arbitrarily shaped void regions from the parent domain: an essential requirement for fluid simulations around complex objects. Carving out objects produces an incomplete octree. We develop efficient lop -down and bottom -up traversal methods to perform finite element computations on incomplete octrees. We validate the framework by (a) showing appropriate convergence analysis and (b) computing the drag coefficient for flow past a sphere for a wide range of Reynolds numbers (O(1-10(6))) encompassing the drag crisis regime. Finally, we deploy the framework on a realistic geometry on a current. project. 10 evaluate COVID-19 transmission risk in classrooms.

关键词： Incomplete octrees matrix-free finite element method distributed memory parallelism

来源：评论

学校读者我要写书评

暂无评论

Reservoir Echo State Network for Classification of Multivariate Time Series 30

Reservoir Echo State Network for Classification of Multivari...

引用

30th International Conference on High Performance Computing, Data, and Analytics (HiPC)

作者： Purkayastha, Basab Bijoy Barma, Shovan Indian Inst Technol Guwahati Dept Phys Gauhati India Indian Inst Informat Technol Guwahati Dept ECE Gauhati India

ISBN: (纸本)9798350383782;9798350383799

Multivariate time series (MTS) classification has been tackled using various methods, including Reservoir Computing (RC), which generates efficient vectorized representations like reservoir state (RS). RS shines when handling extensive classes or training sets but demands longer processing and substantial memory. Addressing this, in this study we present the Parallel Reservoir Echo State Network (PR-ESN), an optimized parallel training and evaluation algorithm rooted in the ESN principle. It leverages both CPU-shared memory and parallel distributed memory architecture to efficiently capture reservoir state's optimal model space representation, addressing computational challenges in MTS analysis. Distinguishing itself from previous works, PR-ESN combines distributed parallel processing at the network level and shared memory multiprocessing at the node level. This results in reduced memory requirements and faster processing, making it a significant contribution to the field. Key features include PR-ESN's distributed training and evaluation, shared memory parallelization, and MSR concatenation for comprehensive analysis of distributed model space representations. Testing on real-world MTS and benchmark ECG data proves PR-ESN-based classifiers achieve superior accuracy and faster processing times with optimal memory usage. Testing on real-world MTS and benchmark ECG data proves PR-ESN-based classifiers achieve superior accuracy and faster processing times with optimal memory usage.

关键词： Shared memory parallelism distributed memory parallelism Echo State Network Reservoir Computing Parallel Computing Multivariate Time Series Classification

来源：评论

学校读者我要写书评

暂无评论

YewPar: Skeletons for Exact Combinatorial Search 20

YewPar: Skeletons for Exact Combinatorial Search

引用

25th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPoPP)

作者： Archibald, Blair Maier, Patrick Stewart, Robert Trinder, Phil Univ Glasgow Glasgow Lanark Scotland Univ Stirling Stirling Scotland Heriot Watt Univ Edinburgh Midlothian Scotland

ISBN: (纸本)9781450368186

Combinatorial search is central to many applications, yet the huge irregular search trees and the need to respect search heuristics make it hard to parallelise. We aim to improve the reuse of intricate parallel search implementations by providing the first general purpose scalable parallel framework for exact combinatorial search, YewPar. We make the following contributions. (1) We present a novel formal model of parallel backtracking search, covering enumeration, decision, and optimisation search. (2) We introduce Lazy Node Generators as a uniform API for search tree generation. (3) We present the design and implementation of 12 widely applicable algorithmic skeletons for tree search on shared and distributed memory architectures. (4) Uniquely in the field we demonstrate how a wide range of parallel search applications can easily be constructed by composing Lazy Node Generators and the search skeletons. (5) We report a systematic performance analysis of all 12 YewPar skeletons on standard instances of 7 search applications, investigating skeleton overheads and scalability up to 255 workers on 17 distributed locations.

关键词： Combinatorial Search Algorithmic Skeletons distributed memory parallelism HPX

来源：评论

学校读者我要写书评

暂无评论

CAFe: Coarray Fortran Extensions for Heterogeneous Computing 30

CAFe: Coarray Fortran Extensions for Heterogeneous Computing

引用

30th IEEE International Parallel and distributed Processing Symposium (IPDPS)

作者： Rasmussen, Craig Sottile, Matthew Rasmussen, Soren Nagle, Dan Dumas, William Univ Oregon Eugene OR 97403 USA NCAR Boulder CO USA Galois Portland OR USA

ISBN: (纸本)9781509036820

Emerging hybrid accelerator architectures are often proposed for inclusion as components in an exascale machine, not only for performance reasons but also to reduce total power consumption. Unfortunately, programmers of these architectures face a daunting and steep learning curve that frequently requires learning a new language (e.g., OpenCL) or adopting a new programming model. Furthermore, the distributed (and frequently multi-level) nature of the memory organization of clusters of these machines provides an additional level of complexity. This paper presents preliminary work examining how Fortran coarray syntax can be extended to provide simpler access to accelerator architectures. This programming model integrates the Partitioned Global Address Space (PGAS) features of Fortran with some of the more task-oriented constructs in OpenMP 4.0 and OpenACC. It also includes the potential for compiler-based transformations targeting the Open Community Runtime (OCR) environment. We demonstrate these CoArray Fortran extensions (CAFe) by implementing a multigrid Laplacian solver and transforming this high-level code to a mixture of standard coarray Fortran and OpenCL kernels.

关键词： distributed memory parallelism domain specific language

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：