[Date Prev][Date Next][Thread Prev][Thread Next] [Search] [Date Index] [Thread Index]

Re: [MacPerl] MacPERL Performace vs. Intel Class systems



>I cannot understand why the 8500 isn't much faster than a Pentium/60 and
>why a Quadra 800 isn't closer to the same speed as a 486/66.

your benchmark code does too much file IO. to some degree, your code 
isn't so much measuring the performance of the processors as it is the 
performance of the hard drives and possibly memory subsystems (DMAs, 
caches, etc.) of the computers you tested. also, considering that the 
FileManager (the part of the MacOS that interfaces with disks) isn't yet 
native and that that Pentium 60 probably has a 60MHz system bus, i'd say 
the 8500 did rather well in your tests. furthermore, i'm not sure how 
file IO buffering is done in Perl on other platforms, but i wouldn't be 
surprised to learn that different platforms use different buffer sizes. i 
would suspect that larger buffers would improve performance. you could 
set $| = 1; to force autoflushing, but then the Mac's performance is 
likely to get *a lot* worse. i noticed that you didn't try your code on a 
BeOS machine. i'd be interested in the results if you get a chance. if 
you want to exercise the video subsystems on those machines you could try 
this :-)

########### PERLTEST.PL ###############
print "Starting\n";
$starttime = time;
        for ($i = 1; $i <= 100000; $i++) {
                print "$i\n";
        }
$stoptime = time;
print ($stoptime - $starttime), " Seconds elapsed time\n";
#######################################


--
Darin S. Morley
morleyd@acm.org
http://www.stow.com/


***** Want to unsubscribe from this list?
***** Send mail with body "unsubscribe" to mac-perl-request@iis.ee.ethz.ch