From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,86616b1931cbdae5
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: Is Ada likely to survive ?
Date: 1997/08/27
Message-ID: <dewar.872715876@merv>#1/1
X-Deja-AN: 268650604
References: <97080410223317@psavax.pwfl.com>
 <01bca387$42ffbce0$18a9f5cd@asip120> <5sstrj$9ci@portal.gmu.edu>
 <3404215f.0@news.uni-ulm.de>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-08-27T00:00:00+00:00
List-Id: <comp.lang.ada>


Joerg says

<<   a) native cc versus gcc (language ANSI C): nearly vanishing difference in
         performance.
   b) native f77 versus cc/gcc: a slight gain in performance for Fortran,
         but only vaguely visible and also probably depending on the special
         problem to be solved. (I did so matrix calculations.) Other problem
         domains may show a different result.
   c) Ada95 (and this was an now rather old GNAT version: 3.01!): if runtime
         checks are disabled by a corresponding pragma the pure Ada version
         was a factor 2 slower than the f77/cc/gcc versions. As some dis-
         cussions on c.l.a. mentioned this might be due to my use of uncon-
         strained arrays, generics and such things. Also did I not check
         whether the algorithm was in best shape for Ada95 but instead just
         recoded the C/f77-Version in Fortran.>>


You skipped the only meaningful comparison which is to compare using
a mature high performance Fortran compiler, f77 does not meet this
criterion (yet :-)