Zdenek Johan, Kapil K. Mathur, S. Lennart Johnsson, and Thomas J. R. Hughes. Finite element methods on the connection machine cm-5 system. Technical report, Thinking Machines Corporation, 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the improvements are not as dramatic (somewhere between 27 and 40 on 16 processors) This is because, during k way refinement only a few vertices get moved; hence, there is limited cache reuse. 6 Related Work Developing parallel graph partitioning algorithms has received a lot of attention [9, 14, 33, 6, 18, 2, 1, 22] due to its extensive applications in many areas. However, most of this work was concentrated on parallelizing algorithms that produce poor quality partitions, such as serial algorithms based on geometric graph partitioning [14, 6] or algorithms that have very high computational requirements, ....

....in many areas. However, most of this work was concentrated on parallelizing algorithms that produce poor quality partitions, such as serial algorithms based on geometric graph partitioning [14, 6] or algorithms that have very high computational requirements, such as spectral bisection [18]. Recently, a number of researchers have developed graph partitioning algorithms that are based on the more powerful and less expensive multilevel graph partitioning algorithms [33, 1, 22] However, with the exception of our earlier work [22] none of these parallel algorithms perform any ....

Zdenek Johan, Kapil K. Mathur, S. Lennart Johnsson, and Thomas J. R. Hughes. Finite element methods on the connection machine cm-5 system. Technical report, Thinking Machines Corporation, 1993.

....for parallel direct factorizations, and taking advantage the aggregate amount of memory available on parallel computers. Significant amount of work has been done in developing parallel algorithms for partitioning unstructured graphs and for producing fill reducing orderings for sparse matrices [2, 5, 8, 7, 12]. Only moderately good speedups have been obtained for parallel formulation of graph partitioning algorithms that use geometric methods [9, 5] despite the fact that geometric partitioning algorithms are inherently easier to parallelize. All parallel formulations presented so far for spectral ....

....for parallel formulation of graph partitioning algorithms that use geometric methods [9, 5] despite the fact that geometric partitioning algorithms are inherently easier to parallelize. All parallel formulations presented so far for spectral partitioning have reported fairly small speedups [2, 1, 12] unless the graph has been distributed to the processors so that certain degree of data locality is achieved [1] In this paper we present a parallel formulation of a graph partitioning and sparse matrix ordering algorithm that is based on a multilevel algorithm we developed recently [14] A key ....

Zdenek Johan, Kapil K. Mathur, S. Lennart Johnsson, and Thomas J. R. Hughes. Finite element methods on the connection machine cm-5 system. Technical report, Thinking Machines Corporation, 1993.

.... better partitions than those provided by spectral partitioning techniques [22] and are generally at least an order of magnitude faster than even the state of the art implementation of spectral techniques [3] Developing parallel graph partitioning algorithms has received a lot of attention [11, 23, 6, 13, 2, 1, 17] due to its extensive applications in many areas. However, most of this work was concentrated on algorithms based on geometric graph partitioning [11, 6] or algorithms that have very high computational requirements, such as spectral bisection [2, 1, 13] Geometric graph partitioning algorithms ....

....received a lot of attention [11, 23, 6, 13, 2, 1, 17] due to its extensive applications in many areas. However, most of this work was concentrated on algorithms based on geometric graph partitioning [11, 6] or algorithms that have very high computational requirements, such as spectral bisection [2, 1, 13]. Geometric graph partitioning algorithms tend to be inherently parallel, but often produce significantly worse partitions compared with the multilevel algorithms. Due to the high computational complexity of the underlying serial # This work was supported by NSF CCR 9423082, by Army Research ....

Zdenek Johan, Kapil K. Mathur, S. Lennart Johnsson, and Thomas J. R. Hughes. Finite element methods on the connection machine cm-5 system. Technical report, Thinking Machines Corporation, 1993.

.... better partitions than those provided by spectral partitioning techniques [29] and are generally at least an order of magnitude faster than even the state of the art implementation of spectral techniques [3] Developing parallel graph partitioning algorithms has received a lot of attention [13, 31, 6, 15, 2, 1, 19] due to its extensive applications in many areas. However, most of this work was concentrated on algorithms based on geometric graph partitioning [13, 6] or algorithms that have very high computational requirements, such as spectral bisection [2, 1, 15] Geometric graph partitioning algorithms ....

....received a lot of attention [13, 31, 6, 15, 2, 1, 19] due to its extensive applications in many areas. However, most of this work was concentrated on algorithms based on geometric graph partitioning [13, 6] or algorithms that have very high computational requirements, such as spectral bisection [2, 1, 15]. Geometric graph partitioning algorithms tend to be inherently parallel, but often produce significantly worse partitions compared with the multilevel algorithms. Due to the high computational complexity of the underlying serial algorithm, parallel spectral bisection algorithms running even on ....

Zdenek Johan, Kapil K. Mathur, S. Lennart Johnsson, and Thomas J. R. Hughes. Finite element methods on the connection machine cm-5 system. Technical report, Thinking Machines Corporation, 1993.

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