Re: Benchmarks: Graphics vs. batch, Linux vs. Windows

[donalson]

Funny this should come up as I log in to ask a question about timing sims.

I have been trying to find a way in Windows to get the system time (not
swarm time) at the beginning and end of a run so that I can subtract them to
get "real" elapsed time.  I want a long run time so that effects of any
transient responces are minimized and don't want to sit in front of the
computer for 30+ minutes with a stop watch.  Is there a library system call
to get the time from the computer?

Cheers,

   Doug

"M. Lang / S. Railsback" wrote:

> Glen says I should post this:
>
> We just implemented a batch mode on our salmon out-migrant simulator
> (demo'd at SwarmFest; or see http://weasel.cnrs.humboldt.edu/~simsys/).
> I have both Linux (Redhat 5.1; XWindows) and Windows NT on my computer,
> so here is a comparison.
>
> The model has two graphics windows: the animation raster and a
> histogram. The fish-agents do a fair, but not huge, amount of
> computation each time step, mainly looking around for the nearest
> hydraulic-node object that tells them their velocity, and then moving to
> a new location. The space is big: 4,000 X 3,000; I ran 500 fish down it.
>
> Simulations were all done on the same machine, Pentium II, 450 mHz with
> 128k of RAM. Hard disk use during the run appeared negligible. Run time
> measurements were made by eyeball and stop watch.
>
> Windows NT batch mode:  1 min 40 s.
> Linux batch mode:       1:40 (no difference expected from windows)
> Windows with graphics:  13:20
> Linux with graphics:    9:50
>
> So- in Windows batch mode runs 8 times faster; 6 times faster in Linux.
> Windows takes 33% longer for graphics mode than Linux.
>
> Steve R.
> --
> address@hidden
> Lang, Railsback & Assoc.
> 250 California Ave., Arcata CA 95521
> 707-822-0453; Fax 822-1868
>
>                   ==================================
>    Swarm-Support is for discussion of the technical details of the day
>    to day usage of Swarm.  For list administration needs (esp.
>    [un]subscribing), please send a message to <address@hidden>
>    with "help" in the body of the message.

--
*********************************************************************
* Doug Donalson                 Office: (805) 893-2962
* Ecology, Evolution,           Home:   (805) 961-4447
* and Marine Biology            email address@hidden
* UC Santa Barbara
* Santa Barbara Ca. 93106
*********************************************************************
*
*   The most exciting phrase to hear in science, the one that
*   heralds new discoveries, is not "EUREKA" (I have found it) but
*   "That's funny ...?"
*
*       Isaac Asimov
*
*********************************************************************



                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.