#### DMCA

## CGMgraph/CGMlib: Implementing and Testing CGM Graph Algorithms on PC Clusters (2003)

### Cached

### Download Links

Venue: | International Journal of High Performance Computing Applications |

Citations: | 24 - 2 self |

### Citations

1351 |
A bridging model for parallel computation
- VALIANT
- 1990
(Show Context)
Citation Context ...the theoretical PRAM model. Following some previous experimental results for the MASPAR and Cray [14, 6, 18, 17, 11], parallel graph algorithms for more “PC cluster like” parallel models like the =-=BSP [19] a-=-nd CGM [7, 8] have been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrated library ofCGM methods for various graph problems including list ranking, Euler tour, c... |

151 |
Synthesis of Parallel Algorithms
- Reif
- 1993
(Show Context)
Citation Context ... balancing, and CGM partitioning. Both libraries are available for download at http://cgm.dehne.net. 1 Introduction Parallel graph algorithms have been extensively studied in the literature (see e.g. =-=[15]-=- for a survey). However, most of the published parallel graph algorithms have traditionally been designed for the theoretical PRAM model. Following some previous experimental results for the MASPAR an... |

120 | Parallel sorting by regular sampling
- Shi, Schaeffer
- 1992
(Show Context)
Citation Context ... Parallel Prefix Sum: calculatePrefixSum (CommObjectList &result, CommObjectList &data). • Parallel Sorting: sort(CommObjectList &data) using the deterministic parallel sample sort methods in [5] an=-=d [16]. -=-• Request System for exchanging data requests between processors: The CGMlib provides methods sendRequests(...) and sendResponses(...) for routing the requests from their senders to their destinatio... |

92 |
Scalable parallel geometric algorithms for coarse grained multicomputers
- Dehne, Fabri, et al.
- 1993
(Show Context)
Citation Context ...al PRAM model. Following some previous experimental results for the MASPAR and Cray [14, 6, 18, 17, 11], parallel graph algorithms for more “PC cluster like” parallel models like the BSP [19] and =-=CGM [7, 8] h-=-ave been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrated library ofCGM methods for various graph problems including list ranking, Euler tour, connected compon... |

62 | Efficient parallel graph algorithms for coarse grained multicomputers and BSP
- Caceres, Dehne, et al.
- 1997
(Show Context)
Citation Context ...Parallel Graph Algorithms Utilizing the CGM Model CGMgraph provides a list ranking method rankTheList(ObjList<Node> &nodes, ...) which implements a randomized method as well as a deterministic method =-=[15, 9]-=-. The input to the list ranking method is a linear linked list where Tthe pointer is stored as the index ofthe next node. CGMgraph also provides a method getEulerTour(ObjList <Vertex> &r, ...) for Eul... |

36 | List ranking and list scan on the Cray C-90
- Reid-Miller
- 1994
(Show Context)
Citation Context ...survey). However, most of the published parallel graph algorithms have traditionally been designed for the theoretical PRAM model. Following some previous experimental results for the MASPAR and Cray =-=[14, 6, 18, 17, 11], para-=-llel graph algorithms for more “PC cluster like” parallel models like the BSP [19] and CGM [7, 8] have been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrate... |

35 |
Optimization of collective communication operations
- Thakur, Rabenseifner, et al.
- 2005
(Show Context)
Citation Context ...rs. For the h-Relation, the array ns is a different array of random integers for each processor. Figure 4 shows the performance of our h-Relation implementation on THOG, CGM1, ULTRA and SUNFIRE. (See =-=Thakur et al. 2005-=- for related studies.) Figure 5 shows the performance of our prefix sum implementation, and Figure 6 shows the performance of our parallel sort implementation. For n = 5,000,000, THOG is about twice a... |

24 | Parallel implementation of algorithms for finding connected components in graphs
- Hsu, Ramachandran, et al.
- 1997
(Show Context)
Citation Context ...survey). However, most of the published parallel graph algorithms have traditionally been designed for the theoretical PRAM model. Following some previous experimental results for the MASPAR and Cray =-=[14, 6, 18, 17, 11], para-=-llel graph algorithms for more “PC cluster like” parallel models like the BSP [19] and CGM [7, 8] have been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrate... |

21 | Ultimate Parallel list ranking
- Sibeyn
- 1999
(Show Context)
Citation Context ...survey). However, most of the published parallel graph algorithms have traditionally been designed for the theoretical PRAM model. Following some previous experimental results for the MASPAR and Cray =-=[14, 6, 18, 17, 11], para-=-llel graph algorithms for more “PC cluster like” parallel models like the BSP [19] and CGM [7, 8] have been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrate... |

16 | Randomized parallel list ranking for distributed memory multiprocessors
- Dehne, Song
- 1997
(Show Context)
Citation Context ... in the range [1, 5], we expect between c and 10c h-Relation operations. Since the deterministic algorithm is more involved and incurs larger constants, c may be around 10 which would imply a range of=-=[10, 100]-=- for the number of h-Relation operations. We measured usually around 20 h-Relation operations. The number is fairly stable, independent of p, which shows again that log p has little influence on the m... |

13 | A note on coarse grained parallel integer sorting. Parallel Processing
- Chan, Dehne
- 1999
(Show Context)
Citation Context ...lities • Parallel Prefix Sum: calculatePrefixSum (CommObjectList &result, CommObjectList &data). • Parallel Sorting: sort(CommObjectList &data) using the deterministic parallel sample sort methods=-= in [5] a-=-nd [16]. • Request System for exchanging data requests between processors: The CGMlib provides methods sendRequests(...) and sendResponses(...) for routing the requests from their senders to their d... |

13 | Communication efficient parallel sorting - Goodrich - 1996 |

7 | Feasability, portability, predictability and efficiency: Four ambitious goals for the design and of parallel coarse grained graph algorithms - Lassous, Gustedt, et al. - 2000 |

6 | Coarse grained parallel maximum matching in convex bipartite graphs
- Bose, Chan, et al.
- 1999
(Show Context)
Citation Context ...nes, the communication time is a small, essentially fixed, portion ofthe total time. The deterministic list ranking requires between c log p and 2c log ph-Relation operations. With log p in the range =-=[1, 5]-=-, we expect between c and 10c h-Relation operations. Since the deterministic algorithm is more involved and incurs larger constants, c may be around 10 which would imply a range of[10, 100] for the nu... |

6 | Experiments with List Ranking on Explicit Multi-Threaded (XMT) Instruction Parallelism
- Dascal, Vishkin
- 1999
(Show Context)
Citation Context |

5 |
Guest editor’s introduction: Special issue on coarse grained parallel algorithms for scientific applications
- Dehne
- 2006
(Show Context)
Citation Context ...al PRAM model. Following some previous experimental results for the MASPAR and Cray [14, 6, 18, 17, 11], parallel graph algorithms for more “PC cluster like” parallel models like the BSP [19] and =-=CGM [7, 8] h-=-ave been presented in [12, 1–4, 9, 10]. In this paper, we present CGMgraph, the first integrated library ofCGM methods for various graph problems including list ranking, Euler tour, connected compon... |

5 |
CGMGRAPH/CGMLIB: Implementing and Testing
- Chan, Dehne, et al.
- 2005
(Show Context)
Citation Context .... The aim of these libraries is to make efficient parallel graph methods available to a wider community of researchers who can utilize them as building blocks for other parallel programming projects (=-=Chan and Dehne 2003-=-). 1 DEPARTMENT OF MATHEMATICS AND COMPUTER SCIENCE FAYETTEVILLE STATE UNIVERSITY FAYETTEVILLE, NC 28301, USA 2 SCHOOL OF ICT GRIFFITH UNIVERSITY NATHAN, QLD 4111, AUSTRALIA (F.DEHNE@GRIFFITH.EDU.AU; ... |

4 | Coarse Grained Parallel Graph Planarity Testing - Caceres, Chan, et al. - 2000 |

3 | Coarse grained parallel algorithms for detecting convex bipartite graphs - Caceres, Chan, et al. |

3 | List Ranking on Meshes
- Sibeyn
- 1998
(Show Context)
Citation Context |

1 |
Predictability and Efficiency : Four Ambitious Goals for the Design and Implementation of Parallel Coarse Grained Graph Algorithms
- Feasability
(Show Context)
Citation Context ...methods as well as many other CGM algorithms: sorting, prefix sum, one to all broadcast, all to one gather, h-Relation, all to all broadcast, array balancing, and CGM partitioning. In comparison with =-=[12]-=-, CGMgraph implements both a randomized as well as a deterministic list ranking method. Our experimental results for randomized list ranking are similar to the ones reported in [12]. Our implementatio... |