Hello sir.

I wanr to ask whether the intel fortran compiler for linux

can be installed on the nvidia tesla machine

# intel compiler in tesla machine

## intel compiler in tesla machine

For more complete information about compiler optimizations, see our Optimization Notice.

Hello sir,

this means that if i install intel fortran compiler on the host machine it will not utilise

the multi cores of tesla.Is there any tool to make it compatible with cuda to use

mkl pardiso on the tesla?

presently i am using the IVF Compiler with mkl for solving linear equation(pardiso).**My system is intel xeon processor (e5520) with 8 cores.**

I need to solve large sparsematrice around 50,00000 size matrice for many iteration.

the system is taking lot of time.

please give some suggestion how to increase the speed. or changing the processor.

any processor where pardiso isefficient?

Is there any other solver faster than pardiso?

or can we attach one more processor to the present system?

does pardiso works on the cluster?

**50,00000** size matrice for many iteration."
Do you mean the input matrices size is 5 000 000?
What mode ( in-core, out-of-core, hybryd) of PARDISO are you using?
What MKL version?
--Gennady

**iparm(18) -the solver will report the numbers of non-zero elements on the factors.**
**--Gennady**

Hello sir,

i am sending the output for size 12lac size ( 1200000) MATRIx

for 50 lac there is problem with my code,we will see it later,

now please tell me how to make this still faster.

in the start of the program the size of the pf usage is around 3.5 GB

and when it enters the pardiso subroutine it increases to 22.5 GB

nonzero= 15346680

solution start

== PARDISO is running in In-Core mode, because iparam(60)=0 ===

=============== PARDISO: solving a symmetric indef. system ===========

ummary PARDISO: ( reorder to reorder )

===============

imes:

=====

Time fulladj: 0.098160 s

Time reorder: 7.023050 s

Time symbfct: 8.933176 s

Time parlist: 0.275664 s

Time malloc : 0.825073 s

Time total : 19.331269 s total - sum: 2.176147 s

tatistics:

==========

< Parallel Direct Factorization with #processors: > 8

< Numerical Factorization with BLAS3 and O(n) synchronization >

< Linear system Ax = b>

#equations: 1200000

#non-zeros in A: 15346680

non-zeros in A (): 0.001066

#right-hand sides: 1

< Factors L and U >

#columns for each panel: 96

#independent subgraphs: 0

#supernodes: 483385

size of largest supernode: 12302

number of nonzeros in L 2403511054

number of nonzeros in U 1

number of nonzeros in L+U 2403511055

Percentage of computed non-zeros for LL^T factorization

0 %

1 %

2 %

3 %

4 %

5 %

6 %

7 %

8 %

9 %

10 %

11 %

12 %

13 %

14 %

15 %

16 %

17 %

18 %

19 %

20 %

21 %

22 %

23 %

24 %

25 %

26 %

27 %

28 %

29 %

30 %

31 %

32 %

33 %

34 %

35 %

36 %

37 %

38 %

39 %

40 %

41 %

42 %

43 %

44 %

45 %

46 %

47 %

48 %

49 %

50 %

51 %

52 %

53 %

54 %

55 %

56 %

57 %

58 %

59 %

60 %

61 %

62 %

63 %

64 %

65 %

66 %

67 %

68 %

69 %

70 %

71 %

72 %

73 %

74 %

75 %

76 %

77 %

78 %

79 %

80 %

81 %

82 %

83 %

84 %

85 %

86 %

87 %

88 %

89 %

90 %

91 %

92 %

93 %

94 %

95 %

96 %

97 %

98 %

99 %

100 %

=============== PARDISO: solving a symmetric indef. system ===========

ummary PARDISO: ( factorize to factorize )

===============

imes:

=====

Time A to LU: 0.000000 s

Time numfct : 776.535926 s

Time malloc : 0.048327 s

Time total : 776.586925 s total - sum: 0.002672 s

tatistics:

==========

< Parallel Direct Factorization with #processors: > 8

< Numerical Factorization with BLAS3 and O(n) synchronization >

< Linear system Ax = b>

#equations: 1200000

#non-zeros in A: 15346680

non-zeros in A (): 0.001066

#right-hand sides: 1

< Factors L and U >

#columns for each panel: 96

#independent subgraphs: 0

#supernodes: 483385

size of largest supernode: 12302

number of nonzeros in L 2403511054

number of nonzeros in U 1

number of nonzeros in L+U 2403511055

gflop for the numerical factorization: 22853.815471

gflop/s for the numerical factorization: 29.430468

=============== PARDISO: solving a symmetric indef. system ===========

ummary PARDISO: ( solve to solve )

===============

imes:

=====

Time solve : 9.663265 s

Time total : 29.857200 s total - sum: 20.193935 s

tatistics:

==========

< Parallel Direct Factorization with #processors: > 8

< Numerical Factorization with BLAS3 and O(n) synchronization >

< Linear system Ax = b>

#equations: 1200000

#non-zeros in A: 15346680

non-zeros in A (): 0.001066

#right-hand sides: 1

< Factors L and U >

#columns for each panel: 96

#independent subgraphs: 0

#supernodes: 483385

size of largest supernode: 12302

number of nonzeros in L 2403511054

number of nonzeros in U 1

number of nonzeros in L+U 2403511055

gflop for the numerical factorization: 22853.815471

gflop/s for the numerical factorization: 29.430468

solution end

please use the Linker Adviser to have the appropriate linking line.