The distributed asci supercomputer (das) site. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
A Data and Task Parallel Image Processing Environment for.. - Nicolescu, Jonker (2001) (6 citations) (Correct)
....10 measured predicted time(microseconds) n Kb message size (2 Figure 5. Performance of point to point communication on DAS tem which consists of a cluster of Pentium Pro 200Mhz PCs with 64Mb RAM running Linux, and connected through Myrinet in a 3D mesh topology, with dimension order routing [1]. As message passing library we have used MPI [14] Figure 5 shows the performance of point to point communication operations and the predicted communication time. The reported time is the minimum time obtained over 20 executions of the same code. It is reasonable to select the minimum value ....
.... = d which minimizes the err image Pseudocode of the multibaseline The results of the data parallel approach are compared with the results obtained using data and task parallelism on a distributed memory system which consists of a cluster of Pentium Pro 200Mhz PCs with 64Mb RAM running Linux [1], and connected through Myrinet in a 3D mesh topology, with dimension order routing. In the task parallel framework we use a special mechanism to register the images on the processors where they are first created. Moreover, each skeleton has associated the task number to which it corresponds. We ....
The distributed asci supercomputer (das) site.
An Economy Grid Architecture for Service-Oriented Grid.. - Buyya, Abramson, Giddy (2001) (3 citations) (Correct)
....motivated by the public good, prizes, fun, fame, or collaborative advantage. This is clearly evident from the construction of private grids (but on volunteer resources) or research test beds such as Distributed.net [9] SETI Home [20] Condor pool [7] DAS (Distributed ASCI Supercomputer) [10], GUSTO [14] and eGrid [11] Even commercial companies such as Entropia, ProcessTree, Popular Power, Mojo Nation, United Devices, and Parabon are trying exploit idle CPU cycles from desktop machines to build a computational Grid [16] These companies are able to develop large scale ....
Distributed ASCI Supercomputer (DAS) - http://www.cs.vu.nl/das/
An Economy Driven Resource Management Architecture for.. - Buyya, Abramson, Giddy (2000) (19 citations) (Correct)
....The motivations or incentives for contributing resources towards building grids, to date, has been driven by public good, prizes, fun, fame, or collaborative advantage. This is clearly evident from the construction of public or research test beds such as, SETI Home [31] Distributed.net [14] DAS [15], and GUSTO[20] The computational resource contributors to these test beds are mostly motivated by the aforementioned reasons. The chances of gaining access to such computational test beds for solving commercial problems are low. Furthermore, contributing resources to a testbed does not guarantee ....
Distributed ASCI Supercomputer (DAS) - http://www.cs.vu.nl/das/
Breaking the Curse of Dynamics by Task Migration: Pilot.. - Overeinder, Sloot (1997) (2 citations) (Correct)
....parallel computer a sustained performance of one Gflop s has been obtained on scientific applications. An interesting initiative that combines both high speed LAN and WAN interconnections in the implementation of a high performance computing platform is the Distributed ASCI Supercomputer(DAS) [4]. Note that ASCI stands for Advanced School for Computing and Imaging a Dutch research school. The DAS is a 136 node wide area distributed system built out of four Myrinet based Pentium Pro clusters. The four clusters are located at four universities: Free University Amsterdam, University of ....
The distributed ASCI supercomputer (DAS). http://www.asci.tudelft.nl/das/ das.shtml.
Satin: Efficient Parallel Divide-and-Conquer in Java - van Nieuwpoort, Kielmann, Bal (2000) (1 citation) Self-citation (Http Vu) (Correct)
....based on Java. Satin (as the name suggests) was inspired by Cilk. In Satin, single threaded Java programs are parallelized by annotating methods that can run in parallel. Our ultimate goal is to use Satin for distributed supercomputing applications on hierarchical wide area clusters (e.g. the DAS [8]) We think that the divide and conquer model will map eciently on such systems, as the model is also hierarchical. In this paper, however, we focus on the implementation of Satin on a single local cluster computer. In contrast to Atlas, Satin is designed as a compiler based system in order to ....
The Distributed ASCI Supercomputer (DAS). http://www.cs.vu.nl/das/.
Satin: Efficient Parallel Divide-and-Conquer in Java - van Nieuwpoort, Kielmann, Bal (2000) (1 citation) Self-citation (Http Vu) (Correct)
....on Java. Satin (as the name suggests) was inspired by MIT s Cilk. In Satin, single threaded Java programs are parallelized by annotating methods that can run in parallel. Our ultimate goal is to use Satin for distributed supercomputing applications on hierarchical wide area clusters (e.g. the DAS [8]) We think that the divide and conquer model will map efficiently on such systems, as the model is also hierarchical. In this paper, however, we focus on the implementation of Satin on a single local cluster computer. In contrast to Atlas, Satin is designed as a compiler based system, in order to ....
The Distributed ASCI Supercomputer (DAS). http://www.cs.vu.nl/das/.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC