Skip to main content

Table 14 CPU and cache energy consumption of parallel DL trace algorithms on Xeon4

From: String correction using the Damerau-Levenshtein distance

A

B

PP_DL_TRACE

PP_LSDL_TRACE

PP_Strip_TRACE

L vs D

S vs D

S vs L

40000

40000

87.39

118.74

84.85

-35.9%

2.9%

28.5%

80000

80000

334.01

449.89

310.34

-34.7%

7.1%

31.0%

120000

120000

2,433.28

1,149.61

684.28

52.8%

71.9%

40.5%

160000

160000

**

2,149.58

1,202.52

  

44.1%

200000

200000

**

3,524.59

1,898.35

  

46.1%

240000

240000

**

5,410.42

2,684.72

  

50.4%

280000

280000

**

7,707.41

3,657.54

  

52.5%

320000

320000

**

10,384.75

4,789.03

  

53.9%

360000

360000

**

14,612.39

6,200.10

  

57.6%

400000

400000

**

16,559.76

7,472.52

  

54.9%