检索结果-内蒙古大学图书馆

Optimal and near-optimal allocation of precedence-constrained tasks to parallel processors: defying the high complexity using effective search techniques

Optimal and near-optimal allocation of precedence-constraine...

引用

International Conference on parallel processing (ICPP)

作者： I. Ahmad Yu-Kwong Kwok Department of Computer Science Hong Kong University of Science and Technology Hong Kong China Parallel Processing Laboratory School of Electrical and Computer Engineering Purdue University West Lafayette IN USA

Obtaining an optimal schedule for a set of precedence-constrained tasks with arbitrary costs is a well-known NP-complete problem. However, optimal solutions are desired in many situations. In this paper we propose search-based algorithms for determining optimal schedules for moderately large problem sizes. The first algorithm which is based on the A* search technique uses a computationally efficient cost function for guiding the search with reduced complexity. We propose a number of state-pruning techniques to reduce the size of the search space. For further lowering the complexity, we parallelize the search. The parallel version is based on reduced interprocessor communication and is guided by static and dynamic load-balancing schemes to evenly distribute the search states to the processors. We also propose an approximate algorithm that guarantees a bounded deviation from the optimal solution but takes considerably shorter time. Based on an extensive experimental evaluation of the algorithms, we conclude that the parallel algorithm with pruning techniques is an efficient scheme for generating optimal solutions for medium to moderately large problems while the approximate algorithm is a useful alternative if slightly degraded solutions are acceptable.

关键词： Optimal scheduling Processor scheduling Computational efficiency Costs Concurrent computing Network topology Scheduling algorithm Computer science Laboratories Multiprocessor interconnection networks

来源：评论

学校读者我要写书评

暂无评论

An efficient local address generation for the block-cyclic distribution 3

An efficient local address generation for the block-cyclic d...

引用

3rd International Conference on Algorithms and Architectures for parallel processing, ICA3PP 1997

作者： Kwon, Oh-Young Kim, Tae-Geun Han, Tack-Don Yang, Sung-Bong Kim, Shin-Dug Distributed Computing Lab. Systems Engineering Research Institute 1 Ueun-dong Yusong-gu Taejon305-333 Korea Republic of Parallel Processing System Laboratory Dept. of Computer Science Yonsei University Seoul120-749 Korea Republic of

ISBN: (纸本)0780342291

In order to generate local addresses for an array section A(l:h:s) with block-cyclic distribution, an efficient compiling method is required. In this paper, two local address generation methods for the block-cyclic distribution are presented. One is a simple local address generation method that is modified from the virtual-block scheme. The other is a linear-time ΔM table construction method. The array elements of A(l:h:s) to be accessed at run-time build up a family of lines. By using the equation of the lines, a ΔM table can be generated in O(k) time. Experimental results show that a simple local address generation method has poor performance but a linear-time ΔM table generation method is faster than other algorithms in ΔM table generation time and access time for 10,000 array elements. © 1997 IEEE.

关键词： Virtual addresses

来源：评论

学校读者我要写书评

暂无评论

An efficient local address generation for the block-cyclic distribution

An efficient local address generation for the block-cyclic d...

引用

IEEE International Conference on Algorithms and Architectures for parallel processing (ICAP)

作者： Oh-Young Kwon Tae-Geun Kim Tack-Don Han Sung-Bong Yang Shin-Dug Kim Distributed Computing Laboratory Systems Engineering Research Institute Taejon South Korea Parallel Processing System Laboratory Department of Computer Science Yonsei University Seoul South Korea

ISBN: (纸本)0780342291

In order to generate local addresses for an array section A(l:h:s) with block-cyclic distribution, an efficient compiling method is required. In this paper, two local address generation methods for the block-cyclic distribution are presented. One is a simple local address generation method that is modified from the virtual-block scheme. The other is a linear-time /spl Delta/M table construction method. The array elements of A(l:h:s) to be accessed at run-time build up a family of lines. By using the equation of the lines, a /spl Delta/M table can be generated in O(k) time. Experimental results show that a simple local address generation method has poor performance but a linear-time /spl Delta/M table generation method is faster than other algorithms in /spl Delta/M table generation time and access time for 10,000 array elements.

关键词： Runtime Equations distributed computing Systems engineering and theory parallel processing Laboratories Computer science parallel languages Upper bound Automata

来源：评论

学校读者我要写书评

暂无评论

parallel AND distributed TLM COMPUTATION WITH SIGNAL-processing FOR ELECTROMAGNETIC-FIELD MODELING

引用

INTERNATIONAL JOURNAL OF NUMERICAL MODELLING-ELECTRONIC NETWORKS DEVICES AND FIELDS 1995年第3-4期8卷 169-185页

作者： SO, PPM ESWARAPPA, C HOEFER, WJR NSERC/MPR Teltech Research Chair in RF Engineering Department of Electrical and Computer Engineering University of Victoria Victoria British Columbia V8W 3P6 Canada Poman So received his B.Sc. degree in computer science and physics from the University of Toronto Toronto Onatario Canada in 1985. He obtained his B.A. Sc. and M.A.Sc. degrees in electrical engineering (summa cum laude) from the University of Ottawa Ottawa Ontario Canada in 1987 and 1989 respectively. Mr So was a research engineer in the University of Ottawa from January 1989 to October 1991 his research interests included CAD techniques of microwave circuits and numerical methods for electromagnetic wave modelling. He specialized in the development of electromagnetic engineering CAD software and successfully implemented a number of electromagnetic wave simulators based on the two-dimensional and three-dimensional transmission-line matrix methods. From August 1990 to February 1991 he accompanied Professor W. J. R. Hoefer to Rome Italy and Sophia Antiopolas France. During that time he implemented a parallel version of the 3D-TLM simulator for the CM-2 Connection Machine using C-Star and X Windows Library. He is a research engineer and a Ph.D. student at the University of Victoria Victoria British Columbia Canada. He ported the 2D- and 3D-TLM simulators to DECmpp 1200 massively parallel computer using MPL C and C++. He also combined OSA90/hope a commercially available microwave CAD program with a 2D-TLM simulation module for geometry optimization using the Datapipe technique of OSA90/hope. Recently he has developed a distributed client-server computing technique for the TLM method this technique could accelerate the TLM simulation by more than an order of magnitude. Channabasappa Eswarappa received the M.Tech. degree in electrical engineering from the Indian Institute of Technology Kanpur India in 1983 and Ph. D. degree from the University of Ottawa Canada in 1990. He worked as an Assistant Executive Engineer and later as

This paper describes the implementation of transmission-line matrix (TLM) method algorithms on a massively parallel computer (DECmpp 12000), the technique of distributed computing in the UNIX environment, and the combination of TLM analysis with Prony's method as well as with autoregressive moving average (ARMA) digital signal processing for electromagnetic field modelling. By combining these advanced computation techniques, typical electromagnetic field modelling of microwave structures by TLM analysis can be accelerated by a few orders of magnitude.

关键词： Mathematical models

来源：评论

学校读者我要写书评

暂无评论

Multiagent fault‐diagnosis expert system in power systems

引用

European Transactions on Electrical Power 1995年第5期5卷 341-347页

作者： Wang, J.P. Trécat, J. Dr. Jianping Wang is a doctor fellow in the Faculte Polytechnique de Mons in Belgium. His current research field is parallel processing applications in power systems. Dr. Wang obtained his BS degree from Southwest Jiaotong University in P.R. Chinain 1982. From 1982 to 1990 he was appointed as an assistant lecturer. lecturer and associate professor respectively in the electrical engineering department of Southwest Jiaotong University. He obtained his Ph.D in the FacultC Polytechnique de Mons in July of 1993. (BD Dolez 3 1 FPMS B-7000 Mons Belgium T +32-65/37 41 87 Fax +32-65/37 43 00) Professor J. Trécat is a director in the power system laboratory in the Faculty Polytechnique de Mons. He obtained his Ph.D in the University of Manchester Institute of Science and Technology(UM1ST) in U.K. in 1970. He has been working in the fields of power system automation artificial intelligence application and computer vision in the Facult6 Polytechnique de Mons since 1970. He is a senior member of IEEE and he has made much contribution both in electrical engineering education and research. (BD Dolez 3 I FPMS B-7000 Mons T +32-651374185. Fax +32-6513743 00)

In order to solve the speed problem and shallow reasoning problem met in current research in fault diagnosis expert system, this paper presents a model based parallel fault diagnosis expert system for energy management systems (EMS). The Petri net model, which corresponds to the relationship between power system components and their protection systems, has been used for the inference in fault diagnosis. A multiagent parallel processing expert system, which is based on five transputers in Sun workstation, has been built for fault diagnosis purposes. The test results show that this expert system is not only fast but also has better performances. Copyright © 1995 John Wiley & Sons, Ltd.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reply

引用

Trends in Neurosciences 1993年第5期16卷 P177-178页

作者： Mitsuo Kawato Hiroaki Gomi ATR Human Information Processing Research Laboratories Kyoto 619-02 Japan Laboratory of Parallel Distributed Processing Research Institute for Electronic Science Hokkaido University Sapporo Hokkaido 060 Japan

来源：评论

学校读者我要写书评

暂无评论

Inverse dynamics of speech motor control 93

Inverse dynamics of speech motor control

引用

Proceedings of the 7th International Conference on Neural Information processing Systems

作者： Makoto Hirayama Eric Vatikiotis-Bateson Mitsuo Kawato ATR Human Information Processing Research Laboratories Kyoto Japan ATR Human Information Processing Research Laboratories Kyoto Japan and Laboratory of Parallel Distributed Processing Research Institute for Electronic Science Hokkaido University Sapporo Hokkaido Japan

Progress has been made in computational implementation of speech production based on physiological data. An inverse dynamics model of the speech articulator's musculo-skeletal system, which is the mapping from articulator trajectories to electromyographic (EMG) signals, was modeled using the acquired forward dynamics model and temporal (smoothness of EMG activation) and range constraints. This inverse dynamics model allows the use of a faster speech motor control scheme, which can be applied to phoneme-to-speech synthesis via musclo-skeletal system dynamics, or to future use in speech recognition. The forward acoustic model, which is the mapping from articulator trajectories to the acoustic parameters, was improved by adding velocity and voicing information inputs to distinguish acoustic parameter differences caused by changes in source characteristics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Physiologically based speech synthesis 92

Physiologically based speech synthesis

引用

Proceedings of the 6th International Conference on Neural Information processing Systems

作者： Makoto Hirayama Eric Vatikiotis-Bateson Kiyoshi Honda Yasuharu Koike Mitsuo Kawato ATR Human Information Processing Research Laboratories Soraku-gun Kyoto Japan ATR Auditory and Visual Perception Research Laboratories ATR Human Information Processing Research Laboratories Soraku-gun Kyoto Japan and Laboratory of Parallel Distributed Processing Research Institute for Electronic Science Hokkaido University Sapporo Hokkaido Japan

ISBN: (纸本)9781558602748

This study demonstrates a paradigm for modeling speech production based on neural networks. Using physiological data from speech utterances, a neural network learns the forward dynamics relating motor commands to muscles and the ensuing articulator behavior that allows articulator trajectories to be generated from motor commands constrained by phoneme input strings and global performance parameters. From these movement trajectories, a second neural network generates PARCOR parameters that are then used to synthesize the speech acoustics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Grid and Cooperative Computing - GCC 2004 Workshops 1

引用

丛书名： Lecture Notes in Computer science

1000年

作者： Hai Jin Yi Pan Nong Xiao Jianhua Sun

ISBN: (数字)9783540302070

ISBN: (纸本)9783540235781

Welcome to the proceedings of GCC2004 and the city of Wuhan. Grid computing has become a mainstream research area in computer science and the GCC conference has become one of the premier forums for presentation of new and exciting research in all aspectsofgridandcooperativecomputing. Theprogramcommitteeispleasedtopresent the proceedings of the 3rd International Conference on Grid and Cooperative Comp- ing (GCC2004), which comprises a collection of excellent technical papers, posters, workshops, and keynote speeches. The papers accepted cover a wide range of exciting topics, including resource grid and service grid, information grid and knowledge grid, grid monitoring,managementand organizationtools, grid portal, grid service, Web s- vices and their QoS, service orchestration, grid middleware and toolkits, software glue technologies, grid security, innovative grid applications, advanced resource reservation andscheduling,performanceevaluationandmodeling,computer-supportedcooperative work, P2P computing, automatic computing, and meta-information management. The conference continues to grow and this year a record total of 581 manuscripts (including workshop submissions) were submitted for consideration. Expecting this growth, the size of the program committee was increased from 50 members for GCC 2003 for 70 in GCC 2004. Relevant differences from previous editions of the conf- ence: it is worth mentioning a signi?cant increase in the number of papers submitted by authors from outside China; and the acceptance rate was much lower than for p- vious GCC conferences. From the 427 papers submitted to the main conference, the program committee selected only 96 regular papers for oral presentation and 62 short papers for poster presentation in the program.

关键词： Theory of Computation Computer Communication Networks Operating Systems Information Systems Applications (incl. Internet) Information Storage and Retrieval User Interfaces and Human Computer Interaction

来源：评论

学校读者我要写书评

暂无评论

Towards Serialization/Deserialization-free State Transfer in Serverless Workflows

引用

ACM Transactions on Computer Systems 1000年

作者： Xingda Wei Fangming Lu Zhuobin Huang Rong Chen Mingyu Wu Haibo Chen Institute of Parallel and Distributed Systems Shanghai Jiao Tong University Shanghai China Institute of Parallel and Distributed Systems Shanghai Jiao Tong University Shanghai China University of Electronic Science and Technology of China Chengdu China Key Laboratory of System Software Chinese Academy of Sciences Beijing China

Serialization and deserialization dominate the state transfer time of serverless workflows, leading to substantial performance penalties when executing various serverless workflow applications. We identify the key reason for serialization and deserialization as a lack of ability to efficiently access the (remote) memory of another function. To this end, we propose RMMap, an OS primitive for remote memory map, which allows a serverless function to directly access the memory of another function, even if it is located remotely. RMMap is the first to completely eliminate serialization and deserialization overhead when transferring states between any pairs of functions in (unmodified) serverless workflows. To make remote memory map efficient and feasible, we co-design it with modern networking (RDMA), OS, language runtime, and serverless platform. Evaluations using real-world serverless workloads show that integrating RMMap with Knative reduces the serverless workflow execution time on Knative by up to 2.6 × and improves resource utilizations by 86.3%.

关键词： Serverless workflow state transfer serialization and deserialization RDMA

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：