检索结果-内蒙古大学图书馆

Fine-grain parallelism using multi-core, Cell/BE, and GPU Systems

PARALLEL computing 2012年第8期38卷 365-390页

作者： [Anonymous] [a] SiPS INESC-ID/IST Universidade Técnica de Lisboa Rua Alves Redol 9 1000-029 Lisbon Portugal [b] CASPER Department of Computer Science University of Cyprus P.O. Box 20537 CY 1678 Nicosia Cyprus [c] The Exelixis Lab Scientific Computing Group Heidelberg Institute for Theoretical Studies Scholss-Wolfsbrunnenweg 35 D-69118 Heidelberg Germany [d] National Center for Supercomputing Applications University of Illinois at Urbana-Champaign 1205 West Clark Street Urbana IL 61801 USA

Currently, we are facing a situation where applications exhibit increasing computational demands and where a large variety of parallel processor systems are available. In this paper we focus on exploiting fine-grain parallelism for three applications with distinct characteristics: a Bioinformatics application (MrBayes), a Molecular Dynamics application (NAMD), and a database application (TPC-H). We assess, side-by-side, the performance of the three applications on general-purpose multi-core processors, the Cell Broadband Engine (Cell/BE), and Graphics Processing Units (GPU). Our results indicate that application performance depends on the characteristics of the parallel architectures and on the computational requirements of the core functions of the respective applications. For MrBayes the best overall performance is achieved on general-purpose multi-core processors, for NAMD on the Cell/BE, and for TPC-H on GPUs. (C) 2011 Elsevier B.V. All rights reserved.

关键词： Multi-core processors Multi-core acelerators Performance evaluation Fine-grain parallelism Scientific workloads Database workloads

来源：评论

学校读者我要写书评

暂无评论

A Combined Dual-stage Framework for Robust Scheduling of Scientific Applications in Heterogeneous Environments with Uncertain Availability

A Combined Dual-stage Framework for Robust Scheduling of Sci...

引用

IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW)

作者： Florina M. Ciorba Timothy Hansen Srishti Srivastava Ioana Banicescu Anthony A. Maciejewski Howard Jay Siegel Center for Information Services and High Performance Computing Technische Universität Dresden Dresden Germany Department of Electrical and Computer Engineering Colorado State University Fort Collins CO USA Department of Computer Science and Engineering Mississippi State University Mississippi MS USA Department of Electrical and Computer Engineering Department of Computer Science Colorado State University Fort Collins CO USA

Scheduling parallel applications on existing or emerging computing platforms is challenging, and, among other attributes, must be efficient and robust. A dual-stage framework is proposed in this paper to evaluate the robustness of efficient resource allocation and dynamic load balancing of scientific applications in heterogeneous computing environments with uncertain availability. The first stage employs robust resource allocation heuristics, while the second stage incorporates robust dynamic loop scheduling techniques. The combined dual-stage framework constitutes a comprehensive framework that enables and provides guarantees for the robust execution of scientific applications in computing systems where uncertainty is caused by various unpredictable perturbations. The paper reports on studies for determining the best techniques to be used for each stage that: (a) maximize the probability that the system make span satisfies a deadline, and (b) minimize the system make span for every given availability level in the system. The usefulness and benefits of the proposed framework are demonstrated via a small scale example.

关键词： Availability Robustness Resource management Uncertainty Runtime Dynamic scheduling Program processors

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous Task Scheduling for Accelerated OpenMP

Heterogeneous Task Scheduling for Accelerated OpenMP

引用

International Symposium on Parallel and Distributed Processing (IPDPS)

作者： Thomas R.W. Scogland Barry Rountree Wu-chun Feng Bronis R. de Supinski Department of Computer Science Virginia Polytechnic Institute and State University Blacksburg VA USA Center for Applied Scientific Computing Lawrence Livemore National Laboratory Livermore CA USA

Heterogeneous systems with CPUs and computational accelerators such as GPUs, FPGAs or the upcoming Intel MIC are becoming mainstream. In these systems, peak performance includes the performance of not just the CPUs but also all available accelerators. In spite of this fact, the majority of programming models for heterogeneous computing focus on only one of these. With the development of Accelerated Open MP for GPUs, both from PGI and Cray, we have a clear path to extend traditional Open MP applications incrementally to use GPUs. The extensions are geared toward switching from CPU parallelism to GPU parallelism. However they do not preserve the former while adding the latter. Thus computational potential is wasted since either the CPU cores or the GPU cores are left idle. Our goal is to create a runtime system that can intelligently divide an accelerated Open MP region across all available resources automatically. This paper presents our proof-of-concept runtime system for dynamic task scheduling across CPUs and GPUs. Further, we motivate the addition of this system into the proposed Open MP for Accelerators standard. Finally, we show that this option can produce as much as a two-fold performance improvement over using either the CPU or GPU alone.

关键词： Graphics processing unit Dynamic scheduling Integrated circuits Acceleration Schedules Programming Runtime

来源：评论

学校读者我要写书评

暂无评论

A multi-level hierarchical index structure for supporting efficient similarity search on tag sets

A multi-level hierarchical index structure for supporting ef...

引用

International Conference on Research Challenges in Information science, RCIS

作者： Jia-Ling Koh Nonhlanhla Shongwe Chung-Wen Cho Department of Computer Science and Information Engineering National Taiwan Normal University Taipei Taiwan Cloud Computing Center for Mobile Applications Industrial Technology and Research Institute Hsinchu Taiwan

Social communication websites has been an emerging type of a Web service that helps users to share their resources. For providing efficient similarity search of tag set in a social tagging system, we propose a multi-level hierarchical index structure to group similar tag sets. Not only the algorithms of similarity searches of tag sets, but also the algorithms of deletion and updating of tag sets by using the constructed index structure are provided. Furthermore, we define a modified hamming distance function on tag sets, which consider the semantically relatedness when comparing the members for evaluating the similarity of two tag sets. This function is more applicable to evaluate the similarity search of two tag sets. A systematic performance study is performed to verify the effectiveness and the efficiency of the proposed strategies. The experiment results show that the proposed MHIB approach further improves the pruning effect of the previous work which constructs a two-level index structure. Especially, the MHIB approach is well scalable with respect to the three parameters when using either the hamming distance or the modified hamming distance for similarity measure. Although the insertion operation of the MHIB approach requires higher cost than the naïve method, with the assistant of the constructed inverted list of clusters, it performs faster than the previous work. Besides, the cost of performing deletion operation by using the MHIB approach is much less than the other two approaches and so is the update operation.

关键词： Tagging Hamming distance Transaction databases Search problems Upper bound Indexing

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

IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW)

作者： Lizhe Wang Jie Tao Holger Marten Achim Streit Samee U. Khan Joanna Kolodziej Dan Chen Center for Earth Observation and Digital Earth Chinese Academy of Sciences China Steinbuch Center for Computing Karlsruhe Institute of Technology Germany Department of Electrical and Computer Engineering North Dakota State University USA Department of Mathematics and Computer Science University of Bielsko-Biała Poland School of Computer China University of Geoscience China

Recently, the computational requirements for large scale 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 GHadoop, 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： Servers Distributed databases Data processing Torque Software computer architecture Computational modeling

来源：评论

学校读者我要写书评

暂无评论

User interactive Internet of things privacy preserved access control

User interactive Internet of things privacy preserved access...

引用

International Conference for Internet Technology and Secured Transactions (ICITST)

作者： Xin Huang Rong Fu Bangdao Chen Tingting Zhang A. W. Roscoe Department of Computer Science Institute for the Future of Computing Oxford Martin School University of Oxford Oxford UK China Information Technology Security Evaluation Center Beijing China Department of Computer Science University of Oxford Oxford UK ITM/IKS Mid Sweden University Sundsvall Sweden

The Internet of Things (IoT) is a network of objects; it is enabled by the Internet technologies. The IoT always collects sensitive data, but inadequate protection may lead to serious user privacy leakage. Thus, privacy protection functions are important to the IoT. Our research aims to provide better privacy protections to IoTs. Firstly, user controlled privacy preserved access control protocol is proposed. Secondly, context aware k-anonymity privacy policy and filter are designed. Thirdly, privacy protection mechanisms are investigated, and the privacy protection mechanisms are improved based on the investigation results. Therefore, users can control which of their personal data is being collected and accessed, who is collecting and accessing such data, and when these are happening.

关键词： Privacy Information filters

来源：评论

学校读者我要写书评

暂无评论

Saliency-Based Fidelity Adaptation Preprocessing for Video Coding

引用

Journal of computer science & Technology 2011年第1期26卷 195-202页

作者：卢少平张松海 Department of Computer Science and Technology Tsinghua University Beijing Engineering Research Center for Intelligent Processing of Visual Media and Content Security

In this paper, we present a video coding scheme which applies the technique of visual saliency computation to adjust image fidelity before compression. To extract visually salient features, we construct a spatio-temporal saliency map by analyzing the video using a combined bottom-up and top-down visual saliency model. We then use an extended bilateral filter, in which the local intensity and spatial scales are adjusted according to visual saliency, to adaptively alter the image fidelity. Our implementation is based on the H.264 video encoder JM12.0. Besides evaluating our scheme with the H.264 reference software, we also compare it to a more traditional foreground-background segmentation-based method and a foveation-based approach which employs Gaussian blurring. Our results show that the proposed algorithm can improve the compression ratio significantly while effectively preserving perceptual visual quality.

关键词： visual saliency bilateral filter fidelity adjustment region-of-interest encoder

来源：评论

学校读者我要写书评

暂无评论

EEG-based emotion recognition in listening music by using support vector machine and linear dynamic system 1

引用

19th International Conference on Neural Information Processing, ICONIP 2012

作者： Duan, Ruo-Nan Wang, Xiao-Wei Lu, Bao-Liang Center for Brain-Like Computing and Machine Intelligence Department of Computer Science and Engineering Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China MOE-Microsoft Key Laboratory for Intelligent Computing and Intelligent Systems Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China Shanghai Key Laboratory of Scalable Computing and Systems Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China MOE Key Laboratory of Systems Biomedicine Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China

ISBN: (数字)9783642344787

ISBN: (纸本)9783642344770

This paper focuses on the variation of EEG at different emotional states. We use pure music segments as stimuli to evoke the exciting or relaxing emotions of subjects. EEG power spectrum is adopted to form features, power spectrum, differential asymmetry, and rational asymmetry. A linear dynamic system approach is applied to smooth the feature sequence. Minimal-redundancy-maximal-relevance algorithm and principal component analysis are used to reduce the dimension of features. We evaluate the performance of support vector machine, k-nearest neighbor classifiers and least-squares classifiers. The accuracy of our proposed method reaches 81.03% on average. And we show that the frequency bands, beta and theta, perform better than other frequency bands in the task of emotion recognition. © 2012 Springer-Verlag.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Modeling the Performance of an Algebraic Multigrid Cycle Using Hybrid MPI/OpenMP

Modeling the Performance of an Algebraic Multigrid Cycle Usi...

引用

International Conference on Parallel Processing (ICPP)

作者： Hormozd Gahvari William Gropp Kirk E. Jordan Martin Schulz Ulrike Meier Yang Department of Computer Science University of Illinois Urbana-Champaign Urbana IL USA Computational Science Center IBM Thomas J. Watson Research Center Cambridge MA USA Center for Applied Scientific Computing Lawrence Livemore National Laboratory Livermore CA USA

The rise of multicore cluster architectures has led to intense interest in using a combination of MPI and OpenMP to more effectively program these machines. We present a performance model for hybrid implementation of the solve cycle of algebraic multigrid (AMG), a popular iterative solver for large sparse linear systems and a key component of many scientific simulations. We validate the model on two leading parallel platforms, and discuss implications for applications programmed in a hybrid model on future machines.

关键词： Bandwidth Multicore processing Message systems Interpolation Computational modeling Instruction sets

来源：评论

学校读者我要写书评

暂无评论

Online vigilance analysis combining video and electrooculography features

Online vigilance analysis combining video and electrooculogr...

引用

19th International Conference on Neural Information Processing, ICONIP 2012

作者： Du, Ruo-Fei Liu, Ren-Jie Wu, Tian-Xiang Lu, Bao-Liang Center for Brain-like Computing and Machine Intelligence Department of Computer Science and Engineering Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China MOE-Microsoft Key Lab. for Intelligent Computing and Intelligent Systems Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China Shanghai Key Laboratory of Scalable Computing and Systems Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China MOE Key Laboratory of Systems Biomedicine Shanghai Jiao Tong University 800 Dongchuan Road Shanghai 200240 China

ISBN: (纸本)9783642344992

In this paper, we propose a novel system to analyze vigilance level combining both video and Electrooculography (EOG) features. For one thing, the video features extracted from an infrared camera include percentage of closure (PERCLOS) and eye blinks, slow eye movement (SEM), rapid eye movement (REM) are also extracted from EOG signals. For another, other features like yawn frequency, body posture and face orientation are extracted from the video by using Active Shape Model (ASM). The results of our experiments indicate that our approach outperforms the existing approaches based on either video or EOG merely. In addition, the prediction offered by our model is in close proximity to the actual error rate of the subject. We firmly believe that this method can be widely applied to prevent accidents like fatigued driving in the future. © 2012 Springer-Verlag.

关键词： Eye movements

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：