检索结果-内蒙古大学图书馆

ieee international symposium on parallel and distributed processing Workshops and Phd Forum 2011年 1577-1578页

作者： Cao, Jiannong Anglano, Cosimo El Baz, Didier Bourgeois, Julien Caron, Eddy Chou, Cheng-Fu Ciaccio, Giuseppe Dell'Amico, Matteo Fuhrmann, Thomas Gargano, Luisa Ghodsi, Ali Huang, Lican Iannello, Giulio Le Fessant, Fabrice Lefevre, Laurent Hong Kong Polytechnic University Hong-Kong Hong Kong Università del Piemonte Orientale A. Avogadro Italy LAAS-CNRS France LIFC University of Franche-Comte France Ecole Normale Superieure de Lyon France National Taiwan University Taiwan Università di Genova Italy EURECOM Sophia Antipolis France Universitat Karlsruhe Germany Università di Salerno Italy UC Berkeley United States Zhejiang Sci.-Tech. University China Università Campus Biomedico Roma Italy INRIA Saclay France INRIA Rhone-Alpes France

来源：评论

学校读者我要写书评

暂无评论

MapReduce Across distributed Clusters for Data-intensive applications

MapReduce Across Distributed Clusters for Data-intensive App...

引用

26th ieee international parallel and distributed processing symposium (IPDPS) / Workshop on High Performance Data Intensive Computing

作者： Wang, Lizhe Tao, Jie Marten, Holger Streit, Achim Khan, Samee U. Kolodziej, Joanna Chen, Dan Chinese Acad Sci Ctr Earth Observat & Digital Earth Beijing 100864 Peoples R China Steinbuch Ctr Comp Karlsruhe Inst Technol Karlsruhe Germany North Dakota State Univ Dept Elect & Comp Engn Fargo ND 58105 USA Univ Bielsko Biala Dept Math & Comp Sci Bielsko Biala Poland China Univ Geosci Sch Comp Beijing Peoples R China

ISBN: (纸本)9780769546766

Recently, the computational requirements for largescale data-intensive analysis of scientific data have grown significantly. In High Energy Physics (HEP) for example, the Large Hadron Collider (LHC) produced 13 petabytes of data in 2010. This huge amount of data are processed on more than 140 computing centers distributed across 34 countries. The MapReduce paradigm has emerged as a highly successful programming model for large-scale data-intensive computing applications. However, current MapReduce implementations are developed to operate on single cluster environments and cannot be leveraged for large-scale distributed data processing across multiple clusters. On the other hand, workflow systems are used for distributed data processing across data centers. It has been reported that the workflow paradigm has some limitations for distributed data processing, such as reliability and efficiency. In this paper, we present the design and implementation of G-Hadoop, a MapReduce framework that aims to enable large-scale distributed computing across multiple clusters. G-Hadoop uses the Gfarm file system as an underlying file system and executes MapReduce tasks across distributed clusters. Experiments of the G-Hadoop framework on distributed clusters show encouraging results.

关键词： MapReduce Hadoop Data Intensive Computing

来源：评论

学校读者我要写书评

暂无评论

A Case Study for an Accelerated DCNN on FPGA-based Embedded distributed System 33

A Case Study for an Accelerated DCNN on FPGA-based Embedded ...

引用

33rd ieee international parallel and distributed processing symposium (IPDPS)

作者： Nestorov, Anna Maria Scolari, Alberto Reggiani, Enrico Stornaiuolo, Luca Santambrogio, Marco D. Politecn Milan Dipartimento Elettron Informaz & Bioingn Milan Italy

ISBN: (纸本)9781538655559

Face Detection (FD) recently became the base of multiple applications requiring low latency but also with limited resources and energy budgets. Deep Convolutional Neural Networks (DCNNs) are especially accurate in FD, but latency requirements and energy budgets call for Field Programmable Gate Arrays (FPGAs)-based solutions, trading flexibility and efficiency. Nonetheless, the offer of FPGAs solutions is limited and different chips often require expensive re-design phases, while developers desire solutions whose resources can scale proportionally to the demands. Therefore, this work presents an FD solution based on a DCNN on a distributed, embedded system with FPGAs, proposing a general approach to reduce the DCNN size and to design its FPGA cores and investigating its accuracy, performance, and energy efficiency.

关键词： DCNN CNN Quantization Embedded FPGA distributed Face detection PYNQ-Z1 HDL HLS

来源：评论

学校读者我要写书评

暂无评论

Unobtrusive Asynchronous Exception Handling with Standard Java Try/Catch Blocks 32

Unobtrusive Asynchronous Exception Handling with Standard Ja...

引用

32nd ieee international parallel and distributed processing symposium (IPDPS)

作者： Mehrabi, Mostafa Giacaman, Nasser Sinnen, Oliver Univ Auckland Dept Elect & Comp Engn Parallel & Reconfigurable Comp Lab Auckland New Zealand

ISBN: (纸本)9781538643686

The sophisticated nature of parallel computing concepts makes parallel programming challenging. This has encouraged higher-level frameworks that conceal much of the complications behind abstraction layers. Paradigms in this category are mostly performance centric, and do not share the same sentiments for the robustness of asynchronous executions. This is while current applications demand consistency in addition to fast performance. Therefore, programming environments that offer high-level support for asynchronous exception handling will have higher chances for popularity. This paper discusses our latest enhancements to @PT, a parallel programming environment that is based on Java annotations. The proposed concept promotes the robustness of parallelized programs by adhering to the familiar exception handling standards of sequential code, and reducing the asynchronous execution concerns at the API level. This study suggests that the concept simplifies efficient management of asynchronous exceptions, which appears to be a challenge in parallel programming.

关键词： parallel programming exception handling asynchronous annotations @PT

来源：评论

学校读者我要写书评

暂无评论

A Framework for Designing and Evaluating distributed Real-Time applications 14

A Framework for Designing and Evaluating Distributed Real-Ti...

引用

18th ieee/ACM international symposium on distributed Simulation and Real Time applications (DS-RT)

作者： Valadares, Arthur Lopes, Cristina Videira Univ Calif Irvine Bren Sch Informat & Comp Sci Irvine CA 92697 USA

ISBN: (纸本)9781479961436

distributed Real-Time (DRT) systems are among the most complex software systems in the field of Computer Science. The contradictory nature of distributed computing and real-time requirements forces prioritization that is dependent on domain-specific knowledge. This combination of distributed and real-time system requirements generates a new set of properties, and requires a new conceptual framework. In this paper, we present a conceptual framework for developing DRT systems. We divide our framework in 3 phases: example applications, design properties, and evaluation. We demonstrate our framework properties with our own distributed Virtual Environment architecture and other 2 examples: Google's Cluster Architecture and collision avoidance systems. Finally we discuss evaluation of DRT systems through the lens of our own architecture, and conclude with general lessons to apply to other DRT systems. Our conceptual framework enables developers of DRT systems to foresee and address issues early in the design. Additionally, we expose general lessons and good practices of evaluation for other DRT systems.

关键词： distributed processing DRT systems Google cluster architecture collision avoidance systems computer science design properties phase distributed computing distributed realtime applications distributed virtual environment architecture domain-specific knowledge evaluation phase example applications phase Fault tolerance Fault tolerant systems Hardware parallel processing Real-time systems Scalability Sensors Fault tolerant systems distributed computing real-time systems dead reckoning tracer computer science distributed processing parallel processing (COMPUTERS) Fault tolerance Scalability conceptual framework Computer hardware Sensors airborne anticollision system Beacon collision avoidance systems Software systems

来源：评论

学校读者我要写书评

暂无评论

An Entertaining Approach to parallel Programming Education 32

An Entertaining Approach to Parallel Programming Education

引用

32nd ieee international parallel and distributed processing symposium (IPDPS)

作者： Buzek, Emanuel Krulis, Martin Charles Univ Prague Fac Math & Phys Parallel Architectures Applicat Algorithms Res Gr Prague Czech Republic

ISBN: (纸本)9781538655559

Despite the facts that multicore CPUs are present in virtually every personal computer or cell phone and distributed systems in the form of cloud services are steadily penetrating various domains of our lives, only a minority of programmers and computer science graduates are able to effectively design and develop parallel and distributed applications. Serial thinking is natural to all humans and it is also encouraged by many computer science curricula. Even though that leading educational institutions are attempting to rectify this trend by introducing parallel programming courses into their study programs, these courses are often dedicated for more experienced students in their fourth of fifth year since mastering modern parallel technologies like OpenMP or CUDA requires certain level of programming skills. It can be argued, that the parallel thinking should be taught much sooner, perhaps even before tertiary education. To this end, we have created an educational platform Parapple that aims to introduce parallelism and related problems like load balancing or synchronization to inexperienced programmers in an entertaining form. Our platform is web-based, so it can run in any modern browser on all operating systems without installation and the users are required to have only a very basic understanding of structural imperative programming.

关键词： parallel programming education simple visual web

来源：评论

学校读者我要写书评

暂无评论

Stratified Sampling for Even Workload Partitioning Applied to IDA* and Delaunay Algorithms 29

Stratified Sampling for Even Workload Partitioning Applied t...

引用

29th ieee international parallel and distributed processing symposium (IPDPS)

作者： Paudel, Jeeva Lelis, Levi H. S. Amaral, Jose Nelson Univ Alberta Dept Comp Sci Edmonton AB T6G 2M7 Canada Univ Fed Vicosa Dept Informat Vicosa MG Brazil

ISBN: (纸本)9781479986484

This work presents Workload Partitioning and Scheduling (WPS), a novel algorithm for evenly partitioning the computational workload of large implicitly-defined worklist-based applications on distributed/shared-memory systems. In WPS, a stratified sampling technique estimates the number of work items that will be processed in each step of the target application. Then WPS uses this estimation to evenly partition and distribute the computational workload. An empirical evaluation on large applications - Iterative-Deepening A* (IDA*) applied to (4x4)- and (5x5) - Sliding-Tile Puzzles, Delaunay Mesh Generation, and Delaunay Mesh Refinement - shows that WPS is applicable to a range of applications. A coordination between WPS and existing work-stealing schedulers for intranode load balancing yields additional speedups in the range of 18% to 40% compared to that achieved with the existing work-stealing schedulers alone. Such a coordination also outperforms an existing workload-partitioning scheme intended specifically for IDA* algorithms by 17% to 36%.

关键词： Delaunay Triangulation IDA Load Balancing Stratified Sampling Workload Partitioning international Development Association mesh generation Load balancing WPS Statistical sampling Workload

来源：评论

学校读者我要写书评

暂无评论

A fault-aware sensor architecture for cooperative mobile applications

A fault-aware sensor architecture for cooperative mobile app...

引用

26th ieee international parallel and distributed processing symposium (IPDPS) / Workshop on High Performance Data Intensive Computing

作者： Kaiser, Joerg Zug, Sebastian Univ Magdeburg Dept Distributed Syst D-39106 Magdeburg Germany

ISBN: (纸本)9780769546766

One of the striking characteristics of mobile embedded systems is the interaction with the physical world. Prerequisite for this interaction is the reliable perception of the environment by sensors. Exploiting remote sensors of an instrumented environment and other mobile systems will extend the range and modalities of sensing and, in principle, will contribute to a better environment perception. However, such as distributed sensor system also puts new substantial challenges on the assessment and dynamic use of sensor data. This paper focus on dependability issues and presents a fault abstraction and fault-handling concept, that encapsulates individual sensor faults and encourages multi-level fault detection. Based on an analysis of sensor faults, our approach generates a validation for each sensor measurement. This allows assessing remote sensor data quality in a uniform way and provides the basis for a distributed mobile application in which new sensor sources can be discovered and exploited dynamically.

关键词： Fault detection Fault abstraction Fault model distributed application Smart sensors

来源：评论

学校读者我要写书评

暂无评论

Topology-aware VM Placement for Network Optimization in Cloud Data Centers 15

Topology-aware VM Placement for Network Optimization in Clou...

引用

15th ieee international symposium on parallel and distributed processing with applications (ISPA) / 16th ieee international Conference on Ubiquitous Computing and Communications (IUCC)

作者： Lian, Zhen Li, Xin Qin, Xiaolin Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing Jiangsu Peoples R China Collaborat Innovat Ctr Novel Software Technol & I Nanjing Jiangsu Peoples R China

ISBN: (纸本)9781538637906

Cloud data centers are hosting more and more complicated applications or services. This makes the network bandwidth becoming essential and critical for the normal application conduction or service provision. Hence, it is necessary to consider the network issue appropriately when deploying virtual machines (VMs) to avoid the network bottleneck and to guarantee the quality of services. In this paper, we focus on the VM placement problem for minimizing the maximal link utilization to avoid network congestion. We formulate the VM placement problem by representing the user requests with resource topologies, and prove the problem to be NP-hard. We present a heuristic algorithm based on the graph theory, which takes the resource topologies into account. The basic idea is to divide the requested VMs into servers with low network communication cost by analyzing the resource topology. We conduct extensive simulations, and the results show that our algorithm has a significant performance improvement on reducing network occupation compared to the best-fit strategy and divide-and-conquer strategy.

关键词： Cloud data center graph theory network optimization subset-sum problem virtual machine placement

来源：评论

学校读者我要写书评

暂无评论

Harnessing the Power of idle GPUs for Acceleration of Biological Sequence Alignment

Harnessing the Power of idle GPUs for Acceleration of Biolog...

引用

23rd ieee international parallel and distributed processing symposium

作者： Ino, Fumihiko Kotani, Yuki Hagihara, Kenichi Osaka Univ Grad Sch Informat Sci & Technol Suita Osaka 5650871 Japan

ISBN: (纸本)9781424437511

This paper-presents a parallel system capable of accelerating biological sequence alignment on the graphics processing unit (GPU) grid. The GPU grid in this paper is a desktop grid system that utilizes idle GPUs and CPUs in the office and home. Our parallel implementation employs a master-worker paradigm to accelerate Liu's OpenGL-based algorithm that runs on a single GPU We integrate this implementation into a screensaver-based grid system. that detects idle resources on which the alignment code can run. We also show some experimental results comparing our implementation with three different implementations running on a single GPU, a single CPU, or multiple CPUs. As a result, we find that a single non-dedicated GPU can provide us almost the same throughput as two dedicated CPUs in our laboratory environment, where GPU-equipped machines are ordinarily used to develop GPU applications.

关键词： Graphics processing unit

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：