In this work we describe a parallel implementation of the Poisson Surface Reconstruction algorithm based on multigrid domain decomposition. We compare implementations using different;models of data-sharing between pro...
详细信息
ISBN:
(数字)9783642103315
ISBN:
(纸本)9783642103308
In this work we describe a parallel implementation of the Poisson Surface Reconstruction algorithm based on multigrid domain decomposition. We compare implementations using different;models of data-sharing between processors and show that a parallel implementation with distributed memory provides the best scalability. Using our method. we are able to parallelize the reconstruction of models from one billion data points on twelve processors across three machines. providing a ninefold speedup in run nit time without sacrificing reconstruction accuracy.
The current trends related to engineering computations in a distributed industrial environment are presented. The key feature is outsourcing the complex numerical computations and making them available to end users vi...
详细信息
ISBN:
(纸本)3540442960
The current trends related to engineering computations in a distributed industrial environment are presented. The key feature is outsourcing the complex numerical computations and making them available to end users via Intranet. The implementation technology is based on the model-driven development of Java components in an environment including Windows-clients and multiprocessor Linux servers running PVM/MPI. A pilot application from electrotechnical industry is presented.
Technological advances in processor power, networking, tetecommunication and multimedia are stimulating the development of applications requiring parallel and distributedcomputing. This new perspective excites the re...
详细信息
ISBN:
(纸本)0780365836
Technological advances in processor power, networking, tetecommunication and multimedia are stimulating the development of applications requiring parallel and distributedcomputing. This new perspective excites the research of new design methodologies that view the software as an "intelligent" collection of agents that interact by coordinating knowledge-based processes. Here we present an actor-based workflow! architecture that would fit naturally to the distributed heterogeneous environments. Actors combine object-oriented and functional programming in order to make the management of concurrency easier for the user.
Thanks to the recent technological advances, a large variety of image data is at our disposal with variable geometric, radiometric and temporal resolution. In many applications the processing of such images needs high...
详细信息
ISBN:
(纸本)9780819492791
Thanks to the recent technological advances, a large variety of image data is at our disposal with variable geometric, radiometric and temporal resolution. In many applications the processing of such images needs high performance computing techniques in order to deliver timely responses e. g. for rapid decisions or real-time actions. Thus, parallel or distributedcomputing methods, Digital Signal Processor (DSP) architectures, Graphical Processing Unit (GPU) programming and Field-Programmable Gate Array (FPGA) devices have become essential tools for the challenging issue of processing large amount of geo-data. The article focuses on the processing and registration of large datasets of terrestrial and aerial images for 3D reconstruction, diagnostic purposes and monitoring of the environment. For the image alignment procedure, sets of corresponding feature points need to be automatically extracted in order to successively compute the geometric transformation that aligns the data. The feature extraction and matching are ones of the most computationally demanding operations in the processing chain thus, a great degree of automation and speed is mandatory. The details of the implemented operations (named LARES) exploiting parallel architectures and GPU are thus presented. The innovative aspects of the implementation are (i) the effectiveness on a large variety of unorganized and complex datasets, (ii) capability to work with high-resolution images and (iii) the speed of the computations. Examples and comparisons with standard CPU processing are also reported and commented.
作者:
Fiolet, VToursel, BUniv Lille 1
Lab Informat Fondamentale Lille CNRS Upresa 8022 F-59655 Villeneuve Dascq France Univ Mons
Serv Informat B-7000 Mons Belgium
The increasing availability of clusters and grids of workstations allows to bring cheap and powerful ressources for distributed datamining. This paper deals with high performance search of association rules. It propos...
详细信息
ISBN:
(数字)9783540319009
ISBN:
(纸本)3540262199
The increasing availability of clusters and grids of workstations allows to bring cheap and powerful ressources for distributed datamining. This paper deals with high performance search of association rules. It proposes to built an "intelligent" database fragmentation and distribution by using a prealable clustering step, a new method called Incremental clustering allows to execute this clustering step in an efficient distributed way.
A distributed constraint optimization problem (DCOP) distributes the variables and constraints among intelligent agents to enable it to be treated as a constraint satisfaction problem. In this paper, we propose a sear...
详细信息
ISBN:
(纸本)9780889866560
A distributed constraint optimization problem (DCOP) distributes the variables and constraints among intelligent agents to enable it to be treated as a constraint satisfaction problem. In this paper, we propose a search method based on the asynchronous distributed optimization (Adopt) algorithm that has been proposed for DCOP. Adopt is a complete method based on a depth-first-search, and operates asynchronously to find the optimal solution for the overall cost. We propose a distributed search method that builds on the merits of Adopt by considering the local load of each agent as opposed to overall optimization. The effect of the proposal was evaluated in simulations.
Enabling high performance, persistent mobile computing has recently become a very active research area. The widespread popularity of mobile computing devices, such as laptops, handheld devices and cell phones, as well...
详细信息
ISBN:
(纸本)0769523129
Enabling high performance, persistent mobile computing has recently become a very active research area. The widespread popularity of mobile computing devices, such as laptops, handheld devices and cell phones, as well as the recent advances in wireless communication technologies are the principal motivators of this research area. However, battery energy limitation is the main challenge towards enabling persistent mobile computing. Several hardware based techniques have been proposed;this has led to more energy-efficient systems. Nevertheless, the problem still remains and there is a consensus that software based techniques have the potential to reduce energy demand and contribute to solve the problem. In this paper, we look into the problem of distributing computational tasks amongst a set of mobile computing devices in a Mobile wireless Ad hoc NETwork (MANET) in such a way that conserves energy and improves performance. In such a distributed environment, the assignment of computational tasks to different devices and the order of their execution play a vital role in energy conservation and performance improvement. The main contributions of this paper are formulating a novel energy-aware scheduling problem and proposing a heuristic algorithm to solve it. Our scheduling algorithm schedules a set of computational tasks, which may have dependencies and communication, into a set of heterogeneous processors in such a way that minimizes both the total consumed energy and the makespan (i.e., the time by which all tasks complete their execution). Experiments show that significant improvement can be achieved by using our scheduler.
Dramatic advances in DNA sequencing technology have made it possible to study microbial environments by direct sequencing of environmental DNA samples. Yet, due to the huge volume and high data complexity, current de ...
详细信息
ISBN:
(纸本)9781450337236
Dramatic advances in DNA sequencing technology have made it possible to study microbial environments by direct sequencing of environmental DNA samples. Yet, due to the huge volume and high data complexity, current de novo assemblers cannot handle large metagenomic datasets or fail to perform assembly with acceptable quality. This paper presents the first parallel solution for decomposing the metagenomic assembly problem without compromising the post-assembly quality. We transform this problem into that of finding weakly connected components in the de Bruijn graph. We propose a novel distributed memory algorithm to identify the connected subgraphs, and present strategies to minimize the communication volume. We demonstrate the scalability of our algorithm on a soil metagenome dataset with 1.8 billion reads. Our approach achieves a runtime of 22 minutes using 1280 Intel Xeon cores for a 421 GB uncompressed FASTQ dataset. Moreover, our solution is generalizable to finding connected components in arbitrary undirected graphs.
This article presents a new distributed approach for generating all prime numbers up to a given limit. From Eratosthenes, who elaborated the first prime sieve (more than 2000 years ago), to the advances of the paralle...
详细信息
This article presents a new distributed approach for generating all prime numbers up to a given limit. From Eratosthenes, who elaborated the first prime sieve (more than 2000 years ago), to the advances of the parallel computers (which have permitted to reach large limits or to obtain the previous results in a shorter time), prime numbers generation still represents an attractive domain of research. Nowadays, prime numbers play a central role in cryptography and their interest has been increased by the very recent proof that primality testing is in P. In this work, we propose a new distributed algorithm which generates all prime numbers in a given finite interval [2,..., n], based on the wheel sieve. As far as we know, this paper designs the first fully distributed wheel sieve algorithm.
The proceedings contain 179 papers. The special focus in this conference is on Personal Computer Based Networks of Workstations, advances in parallel, distributed Computational Models and Video Processing. The topics ...
ISBN:
(纸本)354067442X
The proceedings contain 179 papers. The special focus in this conference is on Personal Computer Based Networks of Workstations, advances in parallel, distributed Computational Models and Video Processing. The topics include: MPI collective operations over IP multicast;an open market-based architecture for distributedcomputing;the multicluster model to the integrated use of multiple workstation clusters;parallel information retrieval on an SCI-based pc-now;a pc-now based parallel extension for a sequential DBMS;the heterogeneous bulk synchronous parallel model;a new computation of shape moments via quadtree decomposition;a java applet to visualize algorithms on reconfigurable mesh;a hardware implementation of pram and its performance evaluation;a non-binary parallel arithmetic architecture;multithreaded parallel computer model with performance evaluation;a high performance microprocessor for multimedia computing;a novel superscalar architecture for fast DCT implementation;computing distance maps efficiently using an optical bus;advanced data layout optimization for multimedia applications;parallel parsing of mpeg video in a multi-threaded multiprocessor environment;parallelization techniques for spatial-temporal occupancy maps from multiple video streams;heuristic solutions for a mapping problem in a TV-anytime server network;a programming environment for real-time parallel vision;parallel low-level image processing on a distributed memory system;congestion-free routing of streaming multimedia content in BMIN-based parallel systems;performance of on-chip multiprocessors for vision tasks;specification techniques for automatic performance analysis tools and controlling distributed shared memory consistency from high level programming languages.
暂无评论