Skip to main content

Table 2 Performance comparison of MCWT-C and MCWT-SG in terms of executing time

From: Massively parallel non-stationary EEG data processing on GPGPU platforms with Morlet continuous wavelet transform

Scale

MCWT-C(s)

MCWT-SG-C(s)

Speedup

MCWT-SG-O(s)

Speedup

Computer#1 (GTX580)

   10

4,513

486

9.3

47

96.0

   20

8,719

702

12.4

77

113.2

   30

12,938

1,024

12.6

108

119.8

   40

17,160

1,568

11.0

136

126.2

   50

21,244

1,924

11.0

167

127.2

   60

25,067

2,419

10.4

198

126.6

   70

29,039

2,871

10.1

230

126.3

Computer#2 (Tesla C2050)

   10

2,342

292

8.0

71

32.9

   20

4,611

570

8.1

144

32.1

   30

6,937

939

7.4

202

34.3

   40

9,357

1,344

7.0

267

35.1

   50

11,654

1,795

6.5

330

35.3

   60

13,991

2,216

6.3

388

36.1

   70

16,328

2,702

6.1

455

35.9

Computer#3 (GTX680)

   10

2,987

228

13.1

39

76.6

   20

5,876

516

11.4

67

87.7

   30

8,753

951

9.2

94

93.1

   40

11,619

1,345

8.6

121

96.1

   50

14,375

1,774

8.1

149

96.5

   60

17,109

2,253

7.6

178

96.1

   70

20,508

2,719

7.5

206

99.6