Skip to main content

Table 5 Results on the speedup for dataset, D1000×1000 for all memory models and partitioning schemes with processors p={2,4,8,16}

From: Parallel and private generalized suffix tree construction and query on genomic data

Method

Distributed

Shared

Hybrid

 

4

8

16

4

8

16

4

8

16

Horizontal

1.19

1.61

2.80

1.11

2.02

3.33

2.31

3.24

4.69

Vertical

1.74

2.31

2.39

2.62

4.45

6.38

3.48

4.66

4.72

Bi-directional

0.77

2.04

2.99

1.02

4.32

5.64

1.37

2.80

4.08