检索结果-内蒙古大学图书馆

作者： Dutta, Sanghamitra Cadambe, Viveck Grover, Pulkit Department of Electrical and Computer Engineering Carnegie Mellon University PittsburghPA15213 United States Department of Electrical Engineering Pennsylvania State University University ParkPA16802 United States

We consider the problem of computing a matrix-vector product Ax using a set of P parallel or distributed processing nodes prone to 'straggling,' i.e., unpredictable delays. Every processing node can access only a fraction ({s}/{N}) of the N -length vector x , and all processing nodes compute an equal number of dot products. We propose a novel error correcting code-that we call 'Short-Dot'-that introduces redundant, shorter dot products such that only a subset of the nodes' outputs are sufficient to compute Ax. To address the problem of straggling in computing matrix-vector products, prior work uses replication or erasure coding to encode parts of the matrix A , but the length of the dot products computed at each processing node is still N. The key novelty in our work is that instead of computing the long dot products as required in the original matrix-vector product, we construct a larger number of redundant and short dot products that only require a fraction of x to be accessed during the computation. Short-Dot is thus useful in a communication-constrained scenario as it allows for only a fraction of x to be accessed by each processing node. Further, we show that in the particular regime where the number of available processing nodes is greater than the total number of dot products, Short-Dot has lower expected computation time under straggling under an exponential model compared to existing strategies, e.g. replication, in a scaling sense. We also derive fundamental limits on the trade-off between the length of the dot products and the recovery threshold, i.e., the required number of processing nodes, showing that Short-Dot is near-optimal. © 1963-2012 IEEE.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

Emerging challenges and opportunities in parallel computing: the cretaceous redux? 09

Emerging challenges and opportunities in parallel computing:...

引用

proceedings of the twenty-first annual symposium on parallelism in algorithms and architectures

作者： Bruce Hendrickson Sandia National Laboratories Albuquerque NM USA

ISBN: (纸本)9781605586069

Despite decades of activity, parallel computing remains immature. Like much of computer science, advances in the field are driven by a mixture of theoretical insights and technological advances. But in parallel computing, the gap between theory and practice remains disconcertingly *** theoretical concepts in parallel computing were developed in the seventies including P-completeness, PRAMs, boolean circuits, and more. A rich set of algorithms and complexity classes was built on top of these models, and these ideas influenced early thinking about parallel machines. But applied parallel computing today is little in uenced by these *** the past twenty years computational scientists and engineers have embraced parallelism as an essential computational tool for solving ever larger and more complex problems. By the mid-nineties, the computational science community had settled on a computing paradigm involving distributed memory machines programmed with explicit message passing. Although not elegant, this approach allows for portability and the use of inexpensive commodity hardware. This hardware is sufficiently different from the abstract models proposed in the seventies that insights from the theory community have had minimal impact on mainstream parallel computing practice.A variety of alternative theoretical models have been proposed in the interim that are more reflective of hardware realities than their counterparts from the seventies. These include the Bulk Synchronous Processing and LogP complexity models as well as computational abstractions like Active Messages partitioned global address space languages. But these concepts and tools have been consigned to niche roles within the larger MPI-dominated parallel computing ***, several forces are rearranging the parallel computing landscape and will raise the importance of ideas and techniques from the theory community. The most familiar change is the rise of multi-core processors. The need for n

关键词： parallel computing

来源：评论

学校读者我要写书评

暂无评论

Untitled

引用

advances IN ENGINEERING SOFTWARE 2011年第5期42卷 217-218页

作者： Ivanyi, P. Topping, B. H. V. Univ Pecs Pecs Hungary Heriot Watt Univ Edinburgh Midlothian Scotland

来源：评论

学校读者我要写书评

暂无评论

Integrating concurrency control and distributed data into workflow frameworks: an actor model perspective

Integrating concurrency control and distributed data into wo...

引用

IEEE International Conference on Systems, Man and Cybernetics

作者： A. Esposito V. Loia Dipartimento di Matematica e Infomatica Università di Salemo Baronissi Salerno Italy

Technological advances in processor power, networking, telecommunications and multimedia are stimulating the development of applications requiring parallel and distributed computing. This new perspective is enticing research into new design methodologies that view the software as an "intelligent" collection of agents that interact by coordinating knowledge-based processes. We present an actor-based workflow architecture that would fit naturally into distributed heterogeneous environments. The actors combine object-oriented and functional programming in order to make the management of concurrency easier for the user.

关键词： Concurrency control Object oriented modeling distributed computing Concurrent computing Message passing Multimedia computing Design methodology Computer architecture Functional programming Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

On Scalability of distributed Simulation and Optimization of Forming Processes

引用

AIP Conference proceedings 2007年第1期908卷 1585-1590页

作者： Frank Thilo Manfred Grauer Rene Menzel Carsten Müller *Information Systems Institute University of Siegen Hölderlinstr. 3 57068 Siegen Germany †GNS mbH Am Gaußberg 2 38114 Braunschweig Germany

advances in the processing power of modern computer hardware and the availability of sophisticated simulation software have fostered the use of virtual prototyping techniques in the automotive industry. However, many engineering problems require simulation and optimization computations so complex that it is imperative to utilize parallel computing to solve these problems in reasonable time. In this paper, FETI‐INDEED — a parallelized simulation code for sheet metal forming — as well as algorithms for distributed optimizations are examined. Because it is anticipated that increasingly large problems will need to be solved in the future, particular focus is put on scalability aspects of these components. Experimental results are provided for optimizations which were performed on a 300 CPU Opteron cluster.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Application of parallel processors to real-time sensor array processing

Application of parallel processors to real-time sensor array...

引用

International Symposium on parallel Processing

作者： D.R. Martinez MIT Lincoln Laboratories USA

Historically, most radar sensor array processing has been implemented using dedicated and specialized processing systems. This approach was necessary because the algorithm computation requirements were several orders of magnitude higher than any commercial supercomputer could provide in an acceptable size, weight and power. Most recently, with the interest by the military to employ commercial-off-the-shelf (COTS) technology, and the rapid advances in computation throughput by COTS supercomputers, we are able to meet the present algorithm requirements with COTS massively parallel processors (MPPs). However, there are still many issues to be overcome to get the most out of a parallel processor for our application. Some of the important issues are the balance between high computation throughput and total exchange communication over the processor network, message passing or shared memory access with high communication throughput for small message sizes, and the scalability of these machines to more demanding algorithm and application requirements. In this paper we present the application requirements, review the important computational kernels, and conclude with our experiences in evaluating representative MPPs.

关键词： Sensor arrays Military computing Throughput Supercomputers Radar Sensor systems Array signal processing Concurrent computing Computer networks Message passing

来源：评论

学校读者我要写书评

暂无评论

Loosely-coupled, mobile replication of objects with transactions

Loosely-coupled, mobile replication of objects with transact...

引用

International Conference on parallel and distributed Systems (ICPADS)

作者： L. Veiga N. Santos R. Lebre P. Ferreira INESC-ID/IST Lisboa Portugal

advances in wireless technology and affordable info-appliances are making mobile computing a reality. However, programmers do have a real hard task while developing mobile distributed applications in which sharing is needed. Such difficulty arises, mainly because programmers are forced to deal with system level issues such as consistency, durability, availability, etc. We designed, implemented and evaluated an object-based platform called M-OBIWAN that releases the programmer from the above mentioned system-level issues. It supports mobile transactions and replication in an integrated way. In contrast with other approaches, M-OBIWAN provides an automatic replication mechanism allowing the creation of dynamic clusters of objects which are accessed within transactions. In addition, the transactional model is adapted to mobile environments. A prototype implementation has been developed. Its performance has been measured with PDAs and desktop machines, linked via Bluetooth.

关键词： Programming profession Mobile computing Personal digital assistants Power system economics Power generation economics Environmental economics Availability Network servers Prototypes Bluetooth

来源：评论

学校读者我要写书评

暂无评论

Noncontiguous I/O accesses through MPI-IO 03

Noncontiguous I/O accesses through MPI-IO

引用

IEEE/ACM International Symposium on Cluster computing and the Grid (CCGRID)

作者： A. Ching A. Choudhary K. Coloma Wei-keng Liao R. Ross W. Gropp Center for Parallel and Distributed Computing Northwestern University Evanston IL USA Argonne National Laboratory Argonne IL USA

ISBN: (纸本)9780769519197

I/O performance remains a weakness of parallel computing systems today. While this weakness is partly attributed to rapid advances in other system components, I/O interfaces available to programmers and the I/O methods supported by file systems have traditionally not matched efficiently with the types of I/O operations that scientific applications perform, particularly noncontiguous accesses. The MPI-IO interface allows for rich descriptions of the I/O patterns desired for scientific applications and implementations such as ROMIO have taken advantage of this ability while remaining limited by underlying file system methods. A method of noncontiguous data access, list I/O, was recently implemented in the parallel Virtual File System (PVFS). We implement support for this interface in the ROMIO MPI-IO implementation. Through a suite of noncontiguous I/O tests we compared ROMIO list I/O to current methods of ROMIO noncontiguous access and found that the list I/O interface provides performance benefits in many noncontiguous cases.

关键词： File systems Testing distributed computing Mathematics Computer science Laboratories parallel processing Programming profession Tiles Checkpointing

来源：评论

学校读者我要写书评

暂无评论

A flexible architecture for remote server-based emulation

A flexible architecture for remote server-based emulation

引用

International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS)

作者： Y. Gu R. Fujimoto College of Computing Georgia Institute of Technology Atlanta GA USA

Network emulation allows one to perform experiments in a controlled laboratory environment by applying network characteristics to live traffic. With recent advances in parallel simulation, it is becoming feasible to emulate large-scale networks at a high level of fidelity. However, high performance computing platforms are required. When such facilities are not locally available, remote computing facilities might be used. This approach is problematic, however, because of the latency introduced in the real time emulation environment to access the remote parallel machine. A distributed server-based architecture is proposed that utilizes local low-fidelity emulators to provide real-time QoS information to distributed applications, coupled with a remote high-fidelity simulator that is used to update and calibrate the local low-fidelity emulators continuously. The accuracy of this server-based emulation system is evaluated using test cases involving a worm propagation scenario.

关键词： Emulation Computer architecture Communication system traffic control Traffic control Computational modeling Large-scale systems High performance computing Delay parallel machines System testing

来源：评论

学校读者我要写书评

暂无评论

Data mining with graphs and matrices

Data mining with graphs and matrices

引用

9th SIAM International Conference on Data Mining 2009, SDM 2009

作者： Wang, Fei Li, Tao Ding, Chris School of Computing and Information Sciences Florida International University United States Department of Computer Science and Engineering University of Texas Arlington TX United States

ISBN: (纸本)9781615671090

The field of data mining increasingly adapt methods and algorithms from advanced matrix computations, graph theory and optimization. Prominent examples are spectral clustering, non-negative matrix factorization, Principal Component Analysis (PCA) and Singular Value Decomposition (SVD), graph-Laplacian based semi-supervised learning, diffusion process, etc. Graph and matrix-based methods are rapidly becoming popular and significant in data mining for the following reasons: (1) Graph and matrix based methods are amenable to vigorous analysis and benefits from the well-established knowledge in matrix computations, graph theory and optimization;(2) Compared to probabilistic and information theoretic approaches, graph and matrix based methods are fast, easy to understand and implement;and (3) Graph and matrix based methods are especially suitable for parallel and distributed-memory computers to solve large scale challenging problems such as searching and extracting patterns from the entire Web. This tutorial will present recent advances in algorithms and methods using graphs and matrices for modeling and analyzing massive, high-dimensional, and nonlinear-structured data. One main goal of the tutorial is to consolidate the recent ideas on data mining using graphs and matrices. We will summarize some open problems contained in this field and propose some future trends. We also wish to attract practitioners who seek novel ideas for applications.

关键词： Data mining

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：