Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!julius.cs.uiuc.edu!apple!sun-barr!newstop!texsun!convex!convex.COM
From: dodson@convex.COM (Dave Dodson)
Newsgroups: comp.lang.fortran
Subject: Re: Fortran vs. C for numerical work (SUMMARY)
Message-ID: <109987@convex.convex.com>
Date: 6 Dec 90 22:12:39 GMT
References: <7200@lanl.gov> <2392:Nov2902:59:0590@kramden.acf.nyu.edu> <72@tdatirv.UUCP> <5760:Dec618:20:2290@kramden.acf.nyu.edu>
Sender: usenet@convex.com
Reply-To: dodson@convex.COM (Dave Dodson)
Organization: Convex Computer Corporation; Richardson, TX
Lines: 19

In article <5760:Dec618:20:2290@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
>In article <72@tdatirv.UUCP> sarima@tdatirv.UUCP (Stanley Friesen) writes:
>> Except for one thing. In scientific computation typical dimensions would be
>> more like 1000 by 1000 by 1000 by 1000 by 50, which requires a great deal
>> more than a mere thirty pointers.
>
>Fine. In your 50-terabyte computation, can you begrudge space for 2050
>pointers so that you save 5% of your run time? I think so.

Wouldn't that be 50+ million pointers?  If the Fortran array is dimensioned
(1000,1000,1000,50) then the C array would be [50][1000][1000][1000], so
you would have 50 pointers pointing into 50,000 pointers pointing into
50,000,000 pointers pointing to the beginnings of linear arrays of length
1000.

----------------------------------------------------------------------

Dave Dodson		                             dodson@convex.COM
Convex Computer Corporation      Richardson, Texas      (214) 497-4234


Brought to you by Super Global Mega Corp .com