stream on SC2000, autopar

From: John D McCalpin (mccalpin@strauss.udel.edu)
Date: Fri Feb 10 1995 - 13:27:10 CST


-rwx------ 1 mccalpin 4000 125 Nov 11 06:56 auto_jobs
-rw------- 1 mccalpin 4000 769 Nov 11 06:49 auto_out.1
-rw------- 1 mccalpin 4000 769 Nov 11 06:49 auto_out.2
-rw------- 1 mccalpin 4000 769 Nov 11 06:50 auto_out.3
-rw------- 1 mccalpin 4000 769 Nov 11 06:50 auto_out.4
-rw------- 1 mccalpin 4000 769 Nov 11 06:50 auto_out.5
-rw------- 1 mccalpin 4000 769 Nov 11 06:50 auto_out.6
-rw------- 1 mccalpin 4000 769 Nov 11 06:51 auto_out.7
-rw------- 1 mccalpin 4000 769 Nov 11 06:51 auto_out.8

#!/bin/csh
foreach i (1 2 3 4 5 6 7 8)
   setenv PARALLEL $i
   echo $i >auto_out.$i
   /bin/time a.out >>& auto_out.$i
end

1
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 411 hundredths of a second
 Wall Time ; time = 411.000 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 35.3461 0.9176 0.9053 0.9352
Scaling : 35.2773 0.9242 0.9071 0.9474
Summing : 36.1307 1.3441 1.3285 1.3660
SAXPYing : 36.7436 1.3219 1.3064 1.3448

real 50.2
user 46.4
sys 2.1
2
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 307 hundredths of a second
 Wall Time ; time = 307.333 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 66.8070 0.4866 0.4790 0.4949
Scaling : 66.7596 0.4906 0.4793 0.5048
Summing : 70.2895 0.6973 0.6829 0.7150
SAXPYing : 72.6663 0.6735 0.6606 0.6814

real 27.2
user 48.7
sys 4.0
3
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 291 hundredths of a second
 Wall Time ; time = 290.833 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 92.6476 0.3554 0.3454 0.3711
Scaling : 92.4535 0.3585 0.3461 0.3780
Summing : 99.0671 0.4939 0.4845 0.5048
SAXPYing : 102.0230 0.4892 0.4705 0.5125

real 20.5
user 52.7
sys 6.8
4
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 327 hundredths of a second
 Wall Time ; time = 327.333 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 112.6705 0.3034 0.2840 0.3909
Scaling : 113.9285 0.2971 0.2809 0.3204
Summing : 121.3994 0.4031 0.3954 0.4304
SAXPYing : 125.3915 0.3929 0.3828 0.4096

real 17.9
user 58.9
sys 8.1
5
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 298 hundredths of a second
 Wall Time ; time = 298.167 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 127.6019 0.2871 0.2508 0.3835
Scaling : 127.6416 0.2721 0.2507 0.3341
Summing : 138.2489 0.3646 0.3472 0.3957
SAXPYing : 142.6928 0.3446 0.3364 0.3573

real 16.3
user 1:04.2
sys 12.2
6
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 315 hundredths of a second
 Wall Time ; time = 314.833 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 107.1941 0.3798 0.2985 0.4616
Scaling : 124.3380 0.3866 0.2574 0.6038
Summing : 147.8766 0.5586 0.3246 0.8794
SAXPYing : 120.3300 0.4706 0.3989 0.7573

real 21.3
user 1:33.0
sys 11.5
7
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 305 hundredths of a second
 Wall Time ; time = 304.667 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 106.3204 0.3869 0.3010 0.4704
Scaling : 111.7248 0.3777 0.2864 0.5106
Summing : 126.3085 0.5441 0.3800 0.7546
SAXPYing : 113.1499 0.4689 0.4242 0.5096

real 21.4
user 1:36.7
sys 12.3
8
--------------------------------------
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLEPRECISION word
--------------------------------------
 CPU time ; time = 315 hundredths of a second
 Wall Time ; time = 315.417 hundredths of a second
 Increase the size of the arrays if this is <30
  and your clock precision is =<1/100 second
 ---------------------------------------------------
Function Max MB/s RMS time Min time Max time
Assignment: 104.5752 0.4578 0.3060 0.6535
Scaling : 99.9894 0.4558 0.3200 0.5444
Summing : 116.4046 0.6224 0.4124 0.8660
SAXPYing : 135.7650 0.5048 0.3536 0.7812

real 23.9
user 1:55.7
sys 12.7



This archive was generated by hypermail 2b29 : Tue Apr 18 2000 - 05:23:04 CDT