Table 7. The CPU Time Consumption (in Minutes) of Each Test Job Concurrently Executed in Duplicate in the zx6000 and K8-32 Systems.

System

 

zx6000

 

K8-32

Compiler

 

efc 7.1

 

ifc 7.1

Arch. opt.a

 

-tpp2 -O3

 

(a)

Library

 

intrinsic BLAS

 

MKL 6.0

 

GOTO 0.7p

 

BLAS

 

ATLAS 3.4.7

Tasking

 

Single

Dup. 1

Dup. 2

 

Single

Dup. 1

Dup. 2

 

Single

Dup. 1

Dup. 2

 

Single

Dup. 1

Dup. 2

 

Single

Dup. 1

Dup. 2

322-1

 

2.18

2.33

2.31

 

2.38

2.40

2.43

 

2.59

2.75

2.73

 

1.90

2.13

2.10

 

2.10

2.17

2.16

322-2

 

58

59

59

 

58

59

59

 

58

59

59

 

49

49

50

 

49

50

49

322-3

 

87

88

88

 

88

89

89

 

88

89

89

 

73

73

73

 

76

74

74

338

 

14

15

14

 

14

14

14

 

14

15

15

 

13

13

13

 

12

12

12

339

 

18

18

18

 

17

18

18

 

17

17

17

 

14

14

14

 

14

14

14

364

 

14

15

14

 

15

15

15

 

15

15

15

 

16

17

17

 

17

17

17

397

 

220

226

226

 

222

228

227

 

223

229

228

 

195

206

219

 

196

197

197

415-1

 

11

12

11

 

11

11

11

 

10

11

11

 

10

10

10

 

9

10

10

415-2

 

22

23

23

 

21

22

22

 

21

21

21

 

23

23

23

 

20

20

20

420-1

 

41

43

43

 

40

42

42

 

39

42

42

 

43

46

49

 

43

45

44

420-2

 

48

52

53

 

48

50

50

 

46

49

50

 

43

47

52

 

43

45

44

424-1

 

32

36

36

 

26

27

27

 

23

24

24

 

41

41

41

 

27

27

28

424-2

 

32

35

36

 

25

27

27

 

23

24

24

 

40

41

41

 

26

28

27

438

 

22

23

23

 

19

19

19

 

18

18

18

 

26

26

26

 

18

18

18

439-1b

 

45(2)

45(2)

45(2)

 

43(2)

43(2)

43(2)

 

43(2)

43(2)

44(2)

 

39(2)

42(2)

54(2)

 

39(2)

39(2)

38(2)

439-2

 

32

32

33

 

32

33

33

 

32

32

33

 

29

28

30

 

30

30

30

447-1

 

70

71

71

 

66

67

67

 

65

65

65

 

71

70

71

 

64

65

65

447-2

 

42

43

43

 

40

40

40

 

37

38

38

 

47

47

47

 

40

41

41

559-4

 

11

12

12

 

11

13

13

 

12

13

13

 

8

8

8

 

8

9

9

560-4

 

13

15

15

 

14

15

15

 

14

15

15

 

9

9

9

 

9

10

10

561-4

 

31

32

32

 

32

33

33

 

33

33

33

 

25

25

26

 

25

26

25

Sum

 

820

852

852

 

801

824

823

 

791

812

813

 

777

795

821

 

728

740

736

a The global options of ifc is set to "-unroll -O2", plus: (a) "-tpp7 -axW -ipo -ipo_obj". b The numbers in the parentheses are the step numbers required in the geometrical optimization calculations; this CPU time is excluded from sum.