Performance of VML Functions

 

Default processor

Pentium® III processor

Pentium® 4 processor

Itanium® processor

float

double

float

double

float

double

float

double

HA

LA

HA

LA

HA

LA

HA

LA

HA

LA

HA

LA

HA

LA

HA

LA

Inv

10.0

10.0

11.2

11.2

10.1

10.1

11.3

11.2

9.9

3.8

11.7

11.7

3.1

3.1

4.3

4.3

Div

14.5

14.5

16.2

16.2

10.9

10.9

16.4

16.4

9.8

4.1

11.8

11.6

4.2

4.2

5.3

5.3

Sqrt

20.2

20.2

34.6

34.6

8.8

8.7

33.4

33.4

8.9

4.2

23.2

17.1

5.2

5.2

7.3

7.3

InvSqrt

21.9

21.9

32.2

32.2

8.1

8.1

31.8

31.8

6.4

3.8

19.5

15.7

5.2

5.2

6.3

6.3

Cbrt

32.1

32.0

45.3

45.3

32.0

32.0

45.4

45.3

22.4

17.3

40.6

32.4

7.2

7.2

10.4

10.4

InvCbrt

31.4

31.4

42.4

42.4

31.3

31.3

42.3

42.3

23.4

15.8

40.8

26.9

7.2

7.2

9.4

9.4

Pow

63.0

63.0

96.4

96.4

62.3

62.3

96.2

96.2

38.9

38.8

85.0

76.0

11.4

11.4

29.2

29.2

Exp

22.3

22.3

37.0

37.0

19.9

19.9

37.0

37.0

10.8

9.2

26.3

20.5

4.2

4.2

6.3

6.3

Ln

20.8

20.8

35.2

35.2

20.9

20.9

37.8

37.8

17.3

12.6

30.0

24.4

7.2

7.2

11.4

11.4

Log10

20.8

20.8

35.1

35.1

20.9

20.9

37.5

37.5

20.1

12.8

32.3

24.3

7.2

7.2

11.4

11.4

Cos

42.8

42.8

54.4

54.4

44.8

24.1

58.4

58.4

21.1

11.8

35.8

35.8

7.2

7.2

8.4

8.4

Sin

38.9

38.9

50.2

50.2

40.5

24.7

50.1

50.1

19.0

10.8

36.7

36.7

6.2

3.1

7.3

7.3

SinCos

66.1

66.1

77.9

77.9

66.3

36.3

78.2

78.2

29.0

18.3

62.3

48.3

8.2

8.2

10.4

10.4

Tan

56.8

56.8

76.6

68.8

59.8

35.0

76.6

68.8

40.2

18.7

66.6

47.8

8.2

8.2

11.4

11.4

Acos

66.2

65.7

97.3

97.3

43.6

36.1

97.3

97.30

25.0

16.2

69.9

51.8

10.2

10.2

15.4

15.4

Asin

64.9

64.9

94.1

94.1

41.7

28.0

94.1

94.1

23.6

15.9

65.5

49.5

10.2

10.2

15.4

15.4

Atan

65.4

65.4

87.7

87.7

38.4

19.8

87.6

87.6

24.5

14.4

75.7

49.3

8.2

8.2

13.5

13.5

Atan2

85.6

85.6

139.7

139.7

86.1

31.5

139.8

139.8

57.3

28.7

139.7

67.3

12.4

12.4

19.8

19.8

Cosh

51.2

51.2

72.0

72.0

35.9

35.9

72.0

72.0

20.5

13.6

36.9

27.1

7.2

7.2

10.4

10.4

Sinh

49.3

49.3

71.4

71.4

40.0

40.0

71.4

71.4

32.7

14.6

39.0

28.8

7.2

7.2

13.3

13.3

Tanh

52.3

52.3

112.1

112.1

52.2

52.2

112.1

112.1

40.8

33.0

69.0

69.0

11.2

11.2

14.4

14.4

Acosh

149.2

149.2

111.4

111.4

43.6

43.6

111.4

111.4

31.3

22.5

88.3

83.1

13.3

13.3

18.5

18.5

Asinh

159.0

159.0

115.9

115.9

45.4

45.4

115.9

115.9

32.9

22.2

104.4

79.6

14.3

14.3

18.5

18.5

Atanh

112.3

112.3

98.0

98.0

36.9

36.9

97.9

97.9

39.0

22.5

92.0

73.6

12.3

12.3

19.5

19.5


Remarks:
1) Units - CPE (Clocks per Element)
2) Data - vectors of 1000 elements with random generated numbers
3) "Default" means X87 code for all IA-32 processors
4) Performance of "default" version was measured on Pentium® III processor

Intel, Pentium, Pentium III, Pentium 4, and Itanium are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries.

Copyright © 2000-2002, Intel Corporation, All Rights Reserved.