## A Comparison of Parallel Algorithms for Connected Components (1994)

Venue: | in the Symposium on Parallel Algorithms and Architectures |

Citations: | 16 - 1 self |

### BibTeX

@INPROCEEDINGS{Greiner94acomparison,

author = {John Greiner},

title = {A Comparison of Parallel Algorithms for Connected Components},

booktitle = {in the Symposium on Parallel Algorithms and Architectures},

year = {1994},

pages = {16--25},

publisher = {Press}

}

### Years of Citing Articles

### OpenURL

### Abstract

This paper presents a comparison of the pragmatic aspects of some parallel algorithms for finding connected components, together with optimizations on these algorithms. The algorithms being compared are two similar algorithms by Shiloach-Vishkin [22] and Awerbuch-Shiloach [2], a randomized contraction algorithm based on algorithms by Reif [21] and Phillips [20], and a hybrid algorithm [11]. Improvements are given for the first two to improve performance significantly, although without improving their asymptotic complexity. The hybrid combines features of the others and is generally the fastest of those tested. Timings were made using NESL [4] code as executed on a Connection Machine 2 and Cray Y-MP/C90. 1 Introduction The complexity of various PRAM algorithms has received much attention, but there has been relatively little work on the implementation and pragmatic efficiency of many of these algorithms. Moreover, much of this work has been for algorithms having regular communication ...

### Citations

255 |
Nonuniversal critical dynamics in Monte Carlo simulations,” Phys
- Swendsen, Wang
- 1987
(Show Context)
Citation Context ...components is important in practice. For example, it is common in computer vision, to group pixels during image analysis; in physics, as part of the Swendsen-Wang algorithm for cluster identification =-=[23]-=-; and VLSI design, for net extraction from circuit masks. Much of the previous pragmatic work has been restricted to the 2D and 3D grids that are of interest in these areas, e.g., [1, 7, 8, 12, 17, 16... |

157 | Scans as primitive parallel operations
- Blelloch
- 1989
(Show Context)
Citation Context ...ith high probability), although for some graphs, including planar graphs, it only requires O(m) work. If scans are assumed to require only constant time, a reasonable assumption for the hardware used =-=[3]-=-, only O(log n) time is needed. There is an optimal, randomizedsO(n)-work, O(log n)-time algorithm by Gazit [9], but it is quite complicated, and for the problem sizes tested here, it is probably not ... |

116 |
An O(log n) parallel connectivity algorithm
- Shiloach, Vishkin
- 1982
(Show Context)
Citation Context ...tic aspects of some parallel algorithms for finding connected components, together with optimizations on these algorithms. The algorithms being compared are two similar algorithms by Shiloach-Vishkin =-=[22]-=- and Awerbuch-Shiloach [2], a randomized contraction algorithm based on algorithms by Reif [21] and Phillips [20], and a hybrid algorithm [11]. Improvements are given for the first two to improve perf... |

95 | NESL: A Nested Data-Parallel Language (version 2.6
- Blelloch
- 1993
(Show Context)
Citation Context ...erformance significantly, although without improving their asymptotic complexity. The hybrid combines features of the others and is generally the fastest of those tested. Timings were made using NESL =-=[4]-=- code as executed on a Connection Machine 2 and Cray Y-MP/C90. 1 Introduction The complexity of various PRAM algorithms has received much attention, but there has been relatively little work on the im... |

47 |
Computing connected components on parallel computers
- Chandra, Sarwate
- 1979
(Show Context)
Citation Context ...ere, it is probably not practical. All of these algorithms are concurrent-read concurrent-write (CRCW). While not investigated here, there has also been numerous exclusivewrite (EW) algorithms, e.g., =-=[6, 13, 14, 15]-=-. Two measures are used for making comparisons. Execution times on a Connection Machine 2 and a Cray YMP /C90 are given for the algorithms, using various sizes and classes of graphs. For the algorithm... |

46 |
An Optimal Randomized Parallel Algorithm for Finding Connected Components in a Graph
- Gazit
- 1991
(Show Context)
Citation Context ...are assumed to require only constant time, a reasonable assumption for the hardware used [3], only O(log n) time is needed. There is an optimal, randomizedsO(n)-work, O(log n)-time algorithm by Gazit =-=[9]-=-, but it is quite complicated, and for the problem sizes tested here, it is probably not practical. All of these algorithms are concurrent-read concurrent-write (CRCW). While not investigated here, th... |

31 |
New connectivity and MSF algorithms for Ultracomputer and PRAM
- Awerbuch, Shiloach
- 1987
(Show Context)
Citation Context ...l algorithms for finding connected components, together with optimizations on these algorithms. The algorithms being compared are two similar algorithms by Shiloach-Vishkin [22] and Awerbuch-Shiloach =-=[2]-=-, a randomized contraction algorithm based on algorithms by Reif [21] and Phillips [20], and a hybrid algorithm [11]. Improvements are given for the first two to improve performance significantly, alt... |

31 | A Comparison of Data-Parallel Algorithms for Connected Components
- Greiner
(Show Context)
Citation Context ...12, 17, 16, 18, 26]. This paper compares implementations and provides optimizations of four algorithms, those of Shiloach-Vishkin (SV) [22], Awerbuch-Shiloach (AS) [2], a "random mate" (RM) =-=algorithm [10]-=-, and a hybrid (HY) of the previous three [11]. The first two algorithms are quite similar and requiresO(m log n) work and O(log n) time. The randomized algorithm combines the random mating of Reif [2... |

29 |
An optimal parallel algorithm for integer sorting
- Reif
- 1985
(Show Context)
Citation Context ...ations on these algorithms. The algorithms being compared are two similar algorithms by Shiloach-Vishkin [22] and Awerbuch-Shiloach [2], a randomized contraction algorithm based on algorithms by Reif =-=[21]-=- and Phillips [20], and a hybrid algorithm [11]. Improvements are given for the first two to improve performance significantly, although without improving their asymptotic complexity. The hybrid combi... |

26 | Fast connected components algorithms for the erew pram
- Karger, Nisan, et al.
- 1999
(Show Context)
Citation Context ...ere, it is probably not practical. All of these algorithms are concurrent-read concurrent-write (CRCW). While not investigated here, there has also been numerous exclusivewrite (EW) algorithms, e.g., =-=[6, 13, 14, 15]-=-. Two measures are used for making comparisons. Execution times on a Connection Machine 2 and a Cray YMP /C90 are given for the algorithms, using various sizes and classes of graphs. For the algorithm... |

21 |
Finding connected components in O(log n log log n) time on the EREW PRAM
- Chong, Lam
- 1995
(Show Context)
Citation Context ...ere, it is probably not practical. All of these algorithms are concurrent-read concurrent-write (CRCW). While not investigated here, there has also been numerous exclusivewrite (EW) algorithms, e.g., =-=[6, 13, 14, 15]-=-. Two measures are used for making comparisons. Execution times on a Connection Machine 2 and a Cray YMP /C90 are given for the algorithms, using various sizes and classes of graphs. For the algorithm... |

15 | Connected components in O(log 3/2 n) parallel time for the CREW
- Johnson, Metaxas
- 1997
(Show Context)
Citation Context |

13 | Hypercube computing: Connected components
- Woo, Sahni
- 1988
(Show Context)
Citation Context ...ever, little work has concentrated on the pragmatic issues involved, and most of the algorithms used in such pragmatic studies have been specialized to grids or specific communication networks, e.g., =-=[19, 25]-=-. The pragmatic issues are of interest, as labeling connected components is important in practice. For example, it is common in computer vision, to group pixels during image analysis; in physics, as p... |

12 |
Parallel graph contraction
- Phillips
- 1989
(Show Context)
Citation Context ...gorithms. The algorithms being compared are two similar algorithms by Shiloach-Vishkin [22] and Awerbuch-Shiloach [2], a randomized contraction algorithm based on algorithms by Reif [21] and Phillips =-=[20]-=-, and a hybrid algorithm [11]. Improvements are given for the first two to improve performance significantly, although without improving their asymptotic complexity. The hybrid combines features of th... |

9 |
A fast parallel algorithm for labeling connected components in image arrays
- Lim, Agrawal, et al.
- 1987
(Show Context)
Citation Context ... identification [23]; and VLSI design, for net extraction from circuit masks. Much of the previous pragmatic work has been restricted to the 2D and 3D grids that are of interest in these areas, e.g., =-=[1, 7, 8, 12, 17, 16, 18, 26]. This pap-=-er compares implementations and provides optimizations of four algorithms, those of Shiloach-Vishkin (SV) [22], Awerbuch-Shiloach (AS) [2], a "random mate" (RM) algorithm [10], and a hybrid ... |

6 | notes: Programming parallel algorithms
- Blelloch, Hardwick
- 1993
(Show Context)
Citation Context ...one edge may be selected pointing from any given node. This restriction is implemented via an implicit concurrent write which arbitrarily picks a single target for the node. The algorithm by Blelloch =-=[5]-=- combines mating with the graph contraction of Phillips [20]. The edges are contracted with the selected, or active, nodes producing supernodes. The edges are contracted by renaming with the new super... |

6 |
Fast algorithms for labeling connected components in 2-D arrays
- Lim
- 1987
(Show Context)
Citation Context ... identification [23]; and VLSI design, for net extraction from circuit masks. Much of the previous pragmatic work has been restricted to the 2D and 3D grids that are of interest in these areas, e.g., =-=[1, 7, 8, 12, 17, 16, 18, 26]. This pap-=-er compares implementations and provides optimizations of four algorithms, those of Shiloach-Vishkin (SV) [22], Awerbuch-Shiloach (AS) [2], a "random mate" (RM) algorithm [10], and a hybrid ... |

4 |
A study of connected component labeling algorithms on the MPP
- Hambrusch, TeWinkel
- 1988
(Show Context)
Citation Context ... identification [23]; and VLSI design, for net extraction from circuit masks. Much of the previous pragmatic work has been restricted to the 2D and 3D grids that are of interest in these areas, e.g., =-=[1, 7, 8, 12, 17, 16, 18, 26]. This pap-=-er compares implementations and provides optimizations of four algorithms, those of Shiloach-Vishkin (SV) [22], Awerbuch-Shiloach (AS) [2], a "random mate" (RM) algorithm [10], and a hybrid ... |

4 |
A Vectorized Algorithm for Cluster Formation in the Swendsen-Wang Dynamics
- Mino
- 1991
(Show Context)
Citation Context |

3 |
A parallel O(log N) algorithm for finding connected components in planar images
- Agrawal, Nekludova, et al.
- 1987
(Show Context)
Citation Context |

3 |
Data-Parallel Connected Components Algorithms
- Greiner, Blelloch
(Show Context)
Citation Context ...g compared are two similar algorithms by Shiloach-Vishkin [22] and Awerbuch-Shiloach [2], a randomized contraction algorithm based on algorithms by Reif [21] and Phillips [20], and a hybrid algorithm =-=[11]-=-. Improvements are given for the first two to improve performance significantly, although without improving their asymptotic complexity. The hybrid combines features of the others and is generally the... |

3 |
An improved algorithm for labeling connected components in a binary image
- Yang
- 1989
(Show Context)
Citation Context |

2 | Vectorized Cluster Search
- Evertz
- 1992
(Show Context)
Citation Context |

2 |
Computational aspects of a parallel algorithm to find the connected components of a graph
- Pardalos, Rentala
- 1989
(Show Context)
Citation Context ...ever, little work has concentrated on the pragmatic issues involved, and most of the algorithms used in such pragmatic studies have been specialized to grids or specific communication networks, e.g., =-=[19, 25]-=-. The pragmatic issues are of interest, as labeling connected components is important in practice. For example, it is common in computer vision, to group pixels during image analysis; in physics, as p... |

1 | Parallel cluster algorithms
- Coddington, Baillie
- 1991
(Show Context)
Citation Context |