X-pop3-spooler: POP3MAIL 2.1.0 b 4 980420 -bs-
Message-ID: <36091A5F.1414DE65@ou.edu>
Date: Wed, 23 Sep 1998 10:57:19 -0500
From: Jim Duchek <jimduchek AT ou DOT edu>
Organization: None
X-Mailer: Mozilla 3.01 (X11; I; Linux 2.0.34 i486)
MIME-Version: 1.0
To: beastium-list AT Desk DOT nl
Subject: AMD K6-2, 3dnow floating point stuff
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: Marc Lehmann <pcg AT goof DOT com>
Status: RO
X-Status: A
Content-Length: 792
Lines: 15

Hi.  I'm not on the mailing list, so if you guys could send the replies
to me, that'd be great.  I just purchased an AMD K6-2 processor. 
Looking at the benchmarks, it's FPU performance is not too shabby
(relative to comparisons between intel/non-intel chips in the past, it's
still pretty far down on real intel).  However, it apparently puts up
Pentium II performance in Quake due to it's floating point SIMD
instructions in it's '3d now' instructions.  I haven't had a chance to
look over the instruction set (and frankly my knowledge dwindles when I
get this close to the processor), but is it possible that these
instructions can be used to optimize FP performance on regular
operations that have nothing to do with 3d graphics?  It seems to me
that they should.

Thanks in advance,
Jim