This is odd to me, and maybe I’m not 100 percent sure what’s going on. But this is the zv6000 again, first running Super Pi under the 32-bit Ubuntu Gnome GUI, and then under the recovery mode console.

This one has Firefox and NFS and a bunch of other stuff still running.

kmandla@cnd5320rkp:~/super_pi$ ./super_pi 20
 Version 2.0 of the super_pi for Linux OS
 Fortran source program was translated into C program with version 19981204 of f2c, then generated C source program was optimized manually.
 pgcc 3.2-3 with compile option of "-fast -tp px -Mbuiltin -Minline=size:1000 -Mnoframe -Mnobounds -Mcache_align -Mdalign -Mnoreentrant" was used for the compilation.
 ------ Started super_pi run : Fri Feb 22 12:12:37 JST 2008
 Start of PI calculation up to 1048576 decimal digits
 End of initialization. Time=       0.996 Sec.
 I= 1 L=       0        Time=       1.948 Sec.
 I= 2 L=       0        Time=       2.128 Sec.
 I= 3 L=       1        Time=       2.144 Sec.
 I= 4 L=       2        Time=       2.136 Sec.
 I= 5 L=       5        Time=       2.128 Sec.
 I= 6 L=      10        Time=       2.132 Sec.
 I= 7 L=      21        Time=       2.132 Sec.
 I= 8 L=      43        Time=       2.132 Sec.
 I= 9 L=      87        Time=       2.128 Sec.
 I=10 L=     174        Time=       2.132 Sec.
 I=11 L=     349        Time=       2.132 Sec.
 I=12 L=     698        Time=       2.132 Sec.
 I=13 L=    1396        Time=       2.148 Sec.
 I=14 L=    2794        Time=       2.148 Sec.
 I=15 L=    5588        Time=       2.128 Sec.
 I=16 L=   11176        Time=       2.108 Sec.
 I=17 L=   22353        Time=       2.092 Sec.
 I=18 L=   44707        Time=       2.044 Sec.
 I=19 L=   89415        Time=       1.916 Sec.
 End of main loop
 End of calculation.    Time=      42.683 Sec.
 End of data output.    Time=       0.188 Sec.
 Total calculation(I/O) time=      42.871(       1.384) Sec.
 ------ Ended super_pi run : Fri Feb 22 12:13:20 JST 2008

Now with background applications closed. A slight bump, and that’s what I expected.

kmandla@cnd5320rkp:~/super_pi$ ./super_pi 20
 Version 2.0 of the super_pi for Linux OS
 Fortran source program was translated into C program with version 19981204 of f2c, then generated C source program was optimized manually.
 pgcc 3.2-3 with compile option of "-fast -tp px -Mbuiltin -Minline=size:1000 -Mnoframe -Mnobounds -Mcache_align -Mdalign -Mnoreentrant" was used for the compilation.
 ------ Started super_pi run : Fri Feb 22 12:16:46 JST 2008
 Start of PI calculation up to 1048576 decimal digits
 End of initialization. Time=       0.948 Sec.
 I= 1 L=       0        Time=       1.828 Sec.
 I= 2 L=       0        Time=       2.100 Sec.
 I= 3 L=       1        Time=       2.100 Sec.
 I= 4 L=       2        Time=       2.100 Sec.
 I= 5 L=       5        Time=       2.100 Sec.
 I= 6 L=      10        Time=       2.104 Sec.
 I= 7 L=      21        Time=       2.096 Sec.
 I= 8 L=      43        Time=       2.100 Sec.
 I= 9 L=      87        Time=       2.100 Sec.
 I=10 L=     174        Time=       2.104 Sec.
 I=11 L=     349        Time=       2.104 Sec.
 I=12 L=     698        Time=       2.096 Sec.
 I=13 L=    1396        Time=       2.100 Sec.
 I=14 L=    2794        Time=       2.104 Sec.
 I=15 L=    5588        Time=       2.100 Sec.
 I=16 L=   11176        Time=       2.088 Sec.
 I=17 L=   22353        Time=       2.060 Sec.
 I=18 L=   44707        Time=       2.016 Sec.
 I=19 L=   89415        Time=       1.880 Sec.
 End of main loop
 End of calculation.    Time=      41.907 Sec.
 End of data output.    Time=       0.184 Sec.
 Total calculation(I/O) time=      42.091(       1.316) Sec.
 ------ Ended super_pi run : Fri Feb 22 12:17:29 JST 2008

Now from the recovery console. Sorry for not copying everything by hand this time.

 End of calculation.    Time=      68.020 Sec.
 End of data output.    Time=       0.352 Sec.
 Total calculation(I/O) time=      68.372(       2.132) Sec.
 ------ Ended super_pi run : Fri Feb 22 12:21:46 JST 2008

Odd, but as you can see, the calculation time was longer under the root console than it was under the Gnome GUI. I doubt very much Gnome was speeding something up, so I can only take the opposite conclusion, that something is bogging down my recovery mode.

I’ll have to double-check how the program works, and what the results really mean. Then maybe play the same trick in 64-bit, and maybe in another distro.