[Fed-Talk] Compilers
[Fed-Talk] Compilers
- Subject: [Fed-Talk] Compilers
- From: Michael Kluskens <email@hidden>
- Date: Mon, 9 May 2005 08:03:34 -0400
I've discovered that a certain unnamed commercial Z compiler can
produce executables that are as much as 2.5x slower than another
compiler for the same Z language (at maximum optimization).
It's very hard to show that a G4 or G5 is faster than the same speed
Pentium with that sort of handicap.
It's also very hard to justify a particular Z language versus C when
dealing that kind of handicap. We're an Z shop not C so I don't have
benchmark C codes (they would have to perform the same functions to be
useable for us).
Now granted the commercial compiler in question is a year or two old,
but I doubt they boosted their speed by 2.5x in one or two years
(without a big statement boasting about it, actually the last patch was
less than a year ago). I purchased their compiler and another
commercial compiler for an Opteron based cluster and didn't see any
such difference.
Now I have a better chance of pushing for a G5 based cluster in the
future once I get certain codes ported to MPI instead of OpenMP (our OS
on the Opteron based cluster corrupted itself and it will take me weeks
to figure out how the vendor configured it). At least format to
reinstall the OS went away in OS X with the release of OS X 10.2
(precisely when I migrated to OS X).
I have not studied the license agreement to determine if posting
comparisons would invalidate my license, etc., so the compiler will
remain unnamed as will the language in question (read about UCITA if
you really don't understand why).
Michael
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Fed-talk mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden