Re: Compiler support for Intel-MMX? (fwd)

Reid Tatge <reid@micro.ti.com>
4 May 1998 22:59:23 -0400

          From comp.compilers
Related articles
Compiler support for Intel-MMX? kyoung@newton.ruph.cornell.edu (1998-04-15)
Re: Compiler support for Intel-MMX? fpeelo@portablesolutions.com (Frank Peelo) (1998-04-18)
Re: Compiler support for Intel-MMX? ba2395@fen.baynet.de (Florian Klaempfl) (1998-04-29)
Re: Compiler support for Intel-MMX? (fwd) reid@micro.ti.com (Reid Tatge) (1998-05-04)
| List of all articles for this month |
From: Reid Tatge <reid@micro.ti.com>
Newsgroups: comp.compilers
Date: 4 May 1998 22:59:23 -0400
Organization: Compilers Central
References: 98-04-068 98-04-070 98-04-101
Keywords: arithmetic, optimize, 386
X-Authentication-Warning: grimm.micro.ti.com: reid owned process doing -bs



Florian Klaempfl wrote:
> ..... If you need _real_ (which is mostly
> important where MMX is used) speed you should use assembler.


I'm sorry, but with statements like that we're going to have to revoke
your license to write compilers. Sure, we all might mumble things
like that among ourselves, after a few beers, when locked in a sealed
room in the basement of an abandoned building and sworn to secrecy
(not that I ever have). But in public we're supposed to present a
united front and claim that that assembly is totally superfluous, an
outdated and crude tool of the unwashed.


Or will be soon... :-)


On a serious note, there is research being done on methods to exploit
"micro-SIMD" instructions, with interesting results. Certainly these
types of instructions present challenges to compilers, but are not
entirely insurmountable.


-Reid
--

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.