台形積分

gistfile1.txt

$ cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 44
model name      : Intel(R) Xeon(R) CPU           L5630  @ 2.13GHz
stepping        : 2
cpu MHz         : 426.069
cache size      : 12288 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu de tsc msr pae cx8 sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm rep_good aperfmperf unfair_spinlock pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 popcnt aes hypervisor lahf_lm ida arat dts
bogomips        : 4266.94
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 44
model name      : Intel(R) Xeon(R) CPU           L5630  @ 2.13GHz
stepping        : 2
cpu MHz         : 426.069
cache size      : 12288 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu de tsc msr pae cx8 sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm rep_good aperfmperf unfair_spinlock pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 popcnt aes hypervisor lahf_lm ida arat dts
bogomips        : 4266.94
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

$ cat /proc/meminfo
MemTotal:        1048576 kB
MemFree:          100672 kB
Cached:           843496 kB
Active:           184796 kB
Inactive:         722624 kB
Active(anon):      11688 kB
Inactive(anon):    52236 kB
Active(file):     173108 kB
Inactive(file):   670388 kB
Unevictable:           0 kB
Mlocked:               0 kB
SwapTotal:       1048576 kB
SwapFree:         456844 kB
Dirty:                 0 kB
Writeback:             0 kB
AnonPages:         63924 kB
Shmem:              2568 kB
Slab:              40468 kB
SReclaimable:      29108 kB
SUnreclaim:        11360 kB

$ gcc --version
gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-4)
Copyright (C) 2010 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.



$ cat log.log
/usr/gridware/sge/default/spool/app01/active_jobs/3119.1/pe_hostfile app01
/usr/gridware/sge/default/spool/app01/active_jobs/3120.1/pe_hostfile app01 /usr/local/pvm3
/var/tmp/aaaNma4HJ
start_pvm: enrolled to local pvmd
start_pvm: got 5 hosts
Warning: no access to tty; thus no job control in this shell...
Sun Microsystems Inc.   SunOS 5.10      Generic January 2005
          =================================================
                    Please Select Kanji Code
          =================================================

                    1.   Shift-JIS

                    2.   EUC

                    3.   UTF-8(default)

                    4.   None

                 Please enter No. =>
a: 0.000000
b: 3.141593
result: 2.000000
result: 0.076120
0.470000[sec]
ログアウト

/usr/gridware/sge/default/spool/app01/active_jobs/3120.1/pe_hostfile app01


$ cat job.sh
#!/bin/sh

#$ -M [email protected]
#$ -m be
#$ -o log.log
#$ -cwd
/usr/local/mpi/bin/mpirun -np $NSLOTS mpi

gistfile2.txt

# 1.1 OpenMP
change parallel
# 1.2 MPI
change parallel
# 2 GridEngine
table n-time-result
program list

mpi.c

#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <time.h>
#include "mpi.h"
#define DIVIDE 10000000


int main(int argc, char **argv){
  clock_t start, stop;
  double a, b;
  double allsum = 0;
  int size, rank;
  MPI_Status status;
  MPI_Init(&argc, &argv);
  MPI_Comm_size(MPI_COMM_WORLD, &size);
  MPI_Comm_rank(MPI_COMM_WORLD, &rank);
  if(rank == 0){
    printf("a: ");printf("%f\n", a=0);
    printf("b: ");printf("%f\n", b=M_PI);
  }
  start = clock();
  double begin = (b-a)/size*rank+a;
  double end = (b-a)/size*(rank+1)+a;
  double step = (b-a)/DIVIDE;
  double sum = 0;
  for(double i=begin; i<=end; i+=step){
    double a = sin(i);
    double b = sin(i+step);
    double h = step;
    double S = (a+b)*h/2;
    sum += S;
  }
  MPI_Reduce(&sum, &allsum, 1, MPI_DOUBLE_PRECISION, MPI_SUM, 0, MPI_COMM_WORLD);
  MPI_Finalize();
  if(rank == 0){
    stop = clock();
    printf("result: %f\n", allsum);
    printf("%f[sec]\n", (double)(stop - start)/CLOCKS_PER_SEC);
  }
  return 0;


/*
$ mpicc -lm -o mpi mpi.c
$ mpirun -np 4 mpi
a: 0.000000
b: 3.141593
result: 2.000000
0.340000[sec]
*/

naive.c

#include <stdio.h>
#include <stdlib.h>
#define M_PI 3.14159265358979323846264338327
#define _USE_MATH_DEFINES
#include <math.h>
#include <time.h>
#define DIVIDE 10000000

int main(int argc, char **argv){
  clock_t start, end;
  double a, b;
  printf("a: ");printf("%f\n", a=0);
  //scanf("%lf", &a);
  printf("b: ");printf("%f\n", b=M_PI);
  //scanf("%lf", &b);
  start = clock();
  double step = (b-a)/DIVIDE;
  double sum = 0;
  for(long i=0; i<=DIVIDE-1; i++){
    double a = sin(step*i);
    double b = sin(step*(i+1));
    double h = step;
    double S = (a+b)*h/2;
    sum += S;
  }
  end = clock();
  printf("result: %f\n", sum);
  printf("%f[sec]\n", (double)(end - start)/CLOCKS_PER_SEC);
  return 0;
}

/*
./native                                                                                                                                                      [/home/legokichi]
a: 0.000000
b: 3.141593
result: 2.000000
0.480000[sec]
*/

openmp.c

#include <stdio.h>
#include <stdlib.h>
#define M_PI 3.14159265358979323846264338327
#define _USE_MATH_DEFINES
#include <math.h>
#include <time.h>
#define DIVIDE 10000000

int main(int argc, char **argv){
  clock_t start, end;
  double a, b;
  printf("a: ");printf("%f\n", a=0);
  //scanf("%lf", &a);
  printf("b: ");printf("%f\n", b=M_PI);
  //scanf("%lf", &b);
  start = clock();
  double step = (b-a)/DIVIDE;
  double sum = 0;
  #pragma omp parallel for reduction(+:sum)
  for(long i=0; i<=DIVIDE-1; i++){
    double a = sin(step*i);
    double b = sin(step*(i+1));
    double h = step;
    double S = (a+b)*h/2;
    sum += S;
  }
  end = clock();
  #ifdef _OPENMP
    printf("use oepnmp\n");
  #endif
  printf("result: %f\n", sum);
  printf("%f[sec]\n", (double)(end - start)/CLOCKS_PER_SEC);
  return 0;
}
/*
$ export OMP_NUM_THREADS=16
$ gcc -std=c99 -lm -fopenmp -O3 openmp.c -o openmp   
$ ./openmp
a: 0.000000
b: 3.141593
use oepnmp
result: 2.000000
0.420000[sec]
*/