Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site watmath.UUCP
Path: utzoo!watmath!wjafyfe
From: wjafyfe@watmath.UUCP (Andy Fyfe)
Newsgroups: net.micro.68k
Subject: Re: Re: FLAME!!! Re: EA orthogonality
Message-ID: <14707@watmath.UUCP>
Date: Tue, 28-May-85 23:57:02 EDT
Article-I.D.: watmath.14707
Posted: Tue May 28 23:57:02 1985
Date-Received: Thu, 30-May-85 01:37:09 EDT
References: <5609@utzoo.UUCP> <75600002@ima.UUCP> <387@gumby.UUCP>
Reply-To: wjafyfe@watmath.UUCP (Andy Fyfe)
Organization: U of Waterloo, Ontario
Lines: 13
Summary: 

In article <387@gumby.UUCP> g-frank@gumby.UUCP writes:
>   The problem is C, not Intel.  If you have programs that require enormous
>data arrays, you picked the wrong processor, didn't you?  Otherwise, you just
>picked the wrong language.

The problem isn't just `C'.  Write a Fortran subroutine that has variable bounds
and the Intel Fortran compiler, not being able to bound the array, will assume
the worst (generating very scary code, particularly if it's a multidimensional
array).  For numerical work the above is very likely ---- a pity, given that
the 8086 family has had floating point hardware for so long.

--Andy Fyfe		...!{decvax, allegra, ihnp4, et. al}!watmath!wjafyfe
			wjafyfe@waterloo.csnet