X-pop3-spooler: POP3MAIL 2.1.0 b 4 980420 -bs-
Message-Id: <199807080913.LAA03546@Cal012054.student.utwente.nl>
X-Mailer: XFMail 1.3 [p0] on Linux
X-Priority: 3 (Normal)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0
In-Reply-To: <19980708045815.20996@cerebro.laendle>
Date: Wed, 08 Jul 1998 11:13:21 +0200 (CEST)
Organization: Badmuts Inc.
From: Derkjan de Haan <j DOT d DOT j DOT dehaan AT student DOT utwente DOT nl>
To: Marc Lehmann <pcg AT goof DOT com>
Subject: RE: please benchmark / MMX #2
Cc: beastium <beastium-list AT Desk DOT nl>
Sender: Marc Lehmann <pcg AT goof DOT com>
Status: RO
X-Status: A
Content-Length: 760
Lines: 34


On 08-Jul-98 Marc Lehmann wrote:
> 
> Ok, I revised the patch a bit.. due to my dumbness, gcc actually generated a
> memory access for every mmx->reg transfer *sigh*

For Intel Pentium-MMX (3*100 MHz):

average of 10 runs on an otherwise inactive system. The numbers are the times in
seconds.
 
mmx:        1.31
nommx:    1.32
mmxonly: 1.30

because of these small differences I decided to test a bigger input file
(static mozilla binary, 6 Mb filesize). Here the picture become clearer:

mmx:        25.34
nommx:    29.30
mmxonly: 25.16


regards,

Derkjan de Haan

----------------------------------
E-Mail: Derkjan de Haan <j DOT d DOT j DOT dehaan AT student DOT utwente DOT nl>
Date: 08-Jul-98
Time: 10:56:13

This message was sent by XFMail
----------------------------------