检索结果-内蒙古大学图书馆

Ad Hoc File Systems for High-Performance computing

Journal of computer science & Technology 2020年第1期35卷 4-26页

作者： AndréBrinkmann Kathryn Mohror Weikuan Yu Philip Carns Toni Cortes Scott A.Klasky Alberto Miranda Franz-Josef Pfreundt Robert B.Ross Marc-AndréVef Zentrum für Datenverarbeitung Johannes Gutenberg University MainzMainz 55128Germany Center for Applied Scientific Computing Lawrence Livermore National LaboratoryLivermoreCA 94550U.S.A Department of Computer Science Florida State UniversityTallahasseeFL 32306U.S.A Mathematics and Computer Science Division Argonne National LaboratoryLemontIL 60439U.S.A Department of Computer Architecture Universitat Politecnica de CatalunyaBarcelona 08034Spain Computer Science and Mathematics Division Oak Ridge National LaboratoryOak RidgeTN 37831U.S.A Computer Science Department Barcelona Supercomputing CenterBarcelona 08034Spain Fraunhofer Institute for Industrial Mathematics ITWM Fraunhofer-Platz 1Kaiserslautern 67663Germany

Storage backends of parallel compute clusters are still based mostly on magnetic disks,while newer and faster storage technologies such as flash-based SSDs or non-volatile random access memory(NVRAM)are deployed within compute *** these new storage technologies into scientific workflows is unfortunately today a mostly manual task,and most scientists therefore do not take advantage of the faster storage *** approach to systematically include nodelocal SSDs or NVRAMs into scientific workflows is to deploy ad hoc file systems over a set of compute nodes,which serve as temporary storage systems for single applications or longer-running *** paper presents results from the Dagstuhl Seminar 17202"Challenges and Opportunities of User-Level File Systems for HPC"and discusses application scenarios as well as design strategies for ad hoc file systems using node-local storage *** discussion includes open research questions,such as how to couple ad hoc file systems with the batch scheduling environment and how to schedule stage-in and stage-out processes of data between the storage backend and the ad hoc file *** presented are strategies to build ad hoc file systems by using reusable components for networking and how to improve storage device *** interfaces and semantics are presented,for example those used by the three ad hoc file systems BeeOND,GekkoFS,and *** presentation covers a range from file systems running in production to cutting-edge research focusing on reaching the performance limits of the underlying devices.

关键词： parallel architectures distributed file system high-performance computing burst buffer POSIX(portable operating system interface)

来源：评论

学校读者我要写书评

暂无评论

Improving gaming experience in zonal MMOGs

Improving gaming experience in zonal MMOGs

引用

15th ACM International Conference on Multimedia, MM'07

作者： Ahmed, Dewan Tanvir Shirmohammadi, Shervin Oliveira, Jauvane DISCOVER Laboratory School of Information Technology and Engineering University of Ottawa Canada ACiMA Laboratory Computer Science Department National Laboratory of Scientific Computing Petrópolis RJ Brazil

ISBN: (纸本)9781595937025

Synchronous communication is a primary concern for multi-user virtual environments like Massively Multiplayer Online Games (MMOGs). Most of the MMOGs offer discrete view for the avatars and follow logical zone layout for easy state management. The avatar movement, from one logical zone to another, causes reorganization at the P2P overlay structure. Its recurrent nature along with unintelligent zone crossing approaches eventually hampers synchronous communication. In this paper, we present performance enhancement mechanisms to reduce P2P overlay reorganization penalties based on avatars' physical characteristics. Avatars' unpredictable movement around the zone boundaries also incur repeated connections and disconnections either among the zone masters or among the multiple peers in the overlay networks. Interest driven zone crossing, dynamic shared region between adjacent zones, and clustering of the entities based on their attributes significantly alleviates these problems and also provides continuous view and seamless region crossing to players. The technical contribution of this paper is to present the effectiveness of object clustering mechanism and interest driven zone crossing with dynamic shared region between each adjacent zones to enhance the gaming experience of MMOG players or users of other types of networked virtual environments (VE). Copyright 2007 ACM.

关键词： Multimedia systems

来源：评论

学校读者我要写书评

暂无评论

DynamicPVM: Dynamic load balancing on parallel systems

DynamicPVM: Dynamic load balancing on parallel systems

引用

International Conference and Exhibition on High-Performance computing and Networking, HPCN 1994

作者： Dikken, Leen van der Linden, Frank Vesseur, Joep Sloot, Peter Shell Nederland Informatieverwerking Rijswijk Netherlands Parallel Scientific Computing and Simulation Group Faculty of Mathematics and Computer Science University of Amsterdam Kruislaan 403 Amsterdam1098 SJ Netherlands

ISBN: (纸本)9783540579816

This paper describes DynamicPVM, an extension to PVM (Parallel Virtual Machine) [1]. PVM enables users to write parallel applications using message passing primitives and statically places the parallel tasks on a collection of nodes. System schedulers schedule atomic jobs over a predefined number of nodes. DynamicPVM addresses the problem of scheduling parallel tasks over a set of nodes. It therefore has to integrate a process checkpointing, migration and restart restart mechanism with the PVM runtime support system. DynamicPVM facilitates an efficient use of existing computational resources for computational jobs consisting of parallel subtasks. Typical target HPC platforms for DynamicPVM are multi user, multi tasking, loosely coupled processors. © 1994, Springer Verlag. All rights reserved.

关键词： Message passing

来源：评论

学校读者我要写书评

暂无评论

Towards detailed material deformation patterns in the hypervelocity impacts of laminated plates with smoothed particle hydrodynamics

Towards detailed material deformation patterns in the hyperv...

引用

ASME 2016 Pressure Vessels and Piping Conference, PVP 2016

作者： Zisis, Iason Van Der Linden, Bas Giannopapa, Christina Koren, Barry Dam, Jacques Centre for Analysis Scientific Computing and Applications Department of Mathematics and Computer Science TU Eindhoven Eindhoven Netherlands Energy Technology Department of Mechanical Engineering TU Eindhoven Eindhoven Netherlands

ISBN: (纸本)9780791850404

Hypervelocity impacts onto the outer structure of spacecraft induce shock waves, which in case of multimaterial structures interact with the material interfaces. This interaction is expected to produce different deformation patterns for laminates compared to the patterns appearing for materials with homogeneous density. In order to exhibit these different features and to show that the Smoothed Particle Hydrodynamics computational method is capable of describing them, a numerical hypervelocity impact experiment is performed for three plates. The first one is a laminate, with discontinuous density and material parameters, the second is a simple homogenized version of the laminate and the last is a monolithic metal plate. Copyright © 2016 by ASME.

关键词： Shock waves

来源：评论

学校读者我要写书评

暂无评论

Scalable compression and replay of communication traces in massively parallel environments

Scalable compression and replay of communication traces in m...

引用

21st International Parallel and Distributed Processing Symposium, IPDPS 2007

作者： Noeth, Michael Mueller, Frank Schulz, Martin De Supinski, Bronis R. North Carolina State University Department of Computer Science Raleigh NC 27695-7534 Lawrence Livermore National Laboratory Center for Applied Scientific Computing Livermore CA 94551

ISBN: (纸本)1424409101

Characterizing the communication behavior of large-scale applications is a difficult and costly task due to code/system complexity and their long execution times. An alternative to running actual codes is to gather their communication traces and then replay them, which facilitates application tuning and future procurements. While past approaches lacked lossless scalable trace collection, we contribute an approach that provides orders of magnitude smaller, if not near constant-size, communication traces regardless of the number of nodes while preserving structural information. We introduce intra- and inter-node compression techniques of MPI events and present results of our implementation for BlueGene/L. Given this novel capability, we discuss its impact on communication tuning and beyond. To the best of our knowledge, such a concise representation of MPI traces in a scalable manner combined with deterministic MPI call replay are without any precedence. ©2007 IEEE.

关键词： Parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

An introduction to using and developing software tools for automatic differentiation

An introduction to using and developing software tools for a...

引用

European Congress on Computational Methods in Applied sciences and Engineering, ECCOMAS 2004

作者： Naumann, Uwe Utke, Jean Walther, Andrea Mathematics and Computer Science Division Argonne National Laboratory Argonne IL 60439 United States Institute for Scientific Computing Technical University Dresden D-01069 Dresden Germany

This paper consists of two parts. Following a brief overview we give a gentle introduction to using various software tools for automatic differentiation. Ready-to-use examples are discussed, and links to further information are presented. Our target audience includes all those who are looking for a straightforward way to get started using the available automatic differentiation technology. The second part of the paper discusses the implementation of source transformation algorithms for automatic differentiation by using an open platform called OpenAD. While the tools presented in the first part have the same mathematical underpinnings, the actual implementations have little in common and mostly are specialized for a particular programming language, compiler internal representation, or purpose. This specialization does not promote an open test bed for experimentation with new algorithms. OpenAD is being designed to fill this need by providing a framework that facilitates the implementation of algorithms that operate on a representation of the numerical kernel of a program. We describe the set of algorithms and basic functionality currently implemented in OpenAD and explain the necessary steps to add a new algorithm to the framework.

关键词： Program compilers

来源：评论

学校读者我要写书评

暂无评论

Pseudo complementary measurement for traditional single-pixel cameras

引用

Chinese Physics B 2020年第11期29卷 315-322页

作者： Qian Dong Xu-Ri Yao Xin Liu Bing Liu Guang-Jie Zhai State Key Laboratory of Computer Science Institute of SoftwareChinese Academy of SciencesBeijing 100190China Key Laboratory of Electronics and Information Technology for Space Systems National Space Science CenterChinese Academy of SciencesBeijing 100190China State Key Laboratory of Scientific and Engineering Computing Academy of Mathematics and Systems ScienceChinese Academy of SciencesBeijing 100190China University of Chinese Academy of Sciences Beijing 100049China

A traditional single-pixel camera needs a large number of measurements to reconstruct the object with compressive sensing *** with the 1/0 matrices in classical measurement,the 1/-1 matrices in the complementary measurement has better property for reconstruction computation and returns better reconstruction ***,each row of the 1/-1 matrices needs two measurements with the traditional single-pixel camera which results into double measurements compared with the 1/0 *** this paper,we consider the pseudo complementary measurement which only takes the same amount of measurements with the row number of some properly designed 1/0 matrix to compute the total luminous flux of the objective and derives the measurement data of the corresponding 1/-1 matrix in a mathematical *** numerical simulation and experimental result show that the pseudo complementary measurement is an efficient tool for the traditional single-pixel camera imaging under low measurement rate,which can combine the advantages of the classical and complementary measurements and significantly improve the peak signal-to-noise ratio.

关键词： computational imaging single-pixel camera

来源：评论

学校读者我要写书评

暂无评论

Two-step iterative methods for solving the stationary convection-diffusion equation with a small parameter at the highest derivative on a uniform grid

引用

Computational Mathematics and Mathematical Physics 2006年第2期46卷 282-293页

作者： Bai, Z.-Z. Krukier, L.A. Martynova, T.S. State Key Laboratory of Scientific/Engineering Computing Institute of Computational Mathematics and Scientific/Engineering Computing Academy of Mathematics and Systems Science Beijing 100080 P.O. Box 2719 China Computer Center Rostov State University Rostov-on-Don 344090 pr. Stachki 200/1 bld. 2 Russian Federation

A stationary convection-diffusion problem with a small parameter multiplying the highest derivative is considered. The problem is discretized on a uniform rectangular grid by the central-difference scheme. A new class of two-step iterative methods for solving this problem is proposed and investigated. The convergence of the methods is proved, optimal iterative methods are chosen, and the rate of convergence is estimated. Numerical results are presented that show the high efficiency of the methods. © MAIK "Nauka/Interperiodica" (Russia), 2006.

关键词： Convergence Dissipative matrix Iterative methods Optimal iterative parameters Transition operator

来源：评论

学校读者我要写书评

暂无评论

Statistical Image Processing: Principle and Algorithms (I)

Statistical Image Processing: Principle and Algorithms (I)

引用

4th International Conference on computer science and Informatics, JCIS 1998

作者： Chen, Li Cooley, Donald H. Zhang, Jianping Scientific and Practical Computing P.O. Box 6081 North Logan UT 84341 Department of Computer Science Utah State University Logan UT 84322-4205

A real digital image processing system contains data acquisition, processing, and display. Image acquisition is to get image data through hardware such as camera and digitizer. Display usually means to show the data on TV monitors or printers. Processing means calculations that are in algorithmic form. Basically, image processing has the following aspects: image transforms, enhancement, restoration, compression and sequential image processing, segmentation, description and registration, reconstruction, and recognition. Statistics plays a very important role for each aspect in general. This paper will introduce several basic statistical methods and algorithms in object search, description and data fitting, segmentation, recognition and interpretation, and sequential image processing. We will also explain how to implement fast statistical algorithms for real image processing.

关键词： Image analysis

来源：评论

学校读者我要写书评

暂无评论

Extending the QUDA library with the eigCG solver 32

Extending the QUDA library with the eigCG solver

引用

32nd International Symposium on Lattice Field Theory, LATTICE 2014

作者： Strelchenko, Alexei Stathopoulos, Andreas Scientific Computing Division Fermilab BataviaIL60510-5011 United States Computer Science Department College of William and Mary WilliamsburgVA23187-8795 United States

While the incremental eigCG algorithm [1] is included in many LQCD software packages, its realization on GPU micro-architectures was still missing. In this session we report our experience of the eigCG implementation in the QUDA library. In particular, we will focus on how to employ the mixed precision technique to accelerate solutions of large sparse linear systems with multiple right-hand sides on GPUs. Although application of mixed precision techniques is a well-known optimization approach for linear solvers, its utilization for the eigenvector computing within eigCG requires special consideration. We will discuss implementation aspects of the mixed precision deflation and illustrate its numerical behavior on the example of the Wilson twisted mass fermion matrix inversions. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence.

关键词： Lattice theory

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：