From owner-freebsd-hardware  Sat Sep 13 12:13:30 1997
Return-Path: <owner-freebsd-hardware>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id MAA21356
          for hardware-outgoing; Sat, 13 Sep 1997 12:13:30 -0700 (PDT)
Received: from gw.rinet.ru (gw.rinet.ru [195.54.192.33])
          by hub.freebsd.org (8.8.7/8.8.7) with SMTP id MAA21351
          for <freebsd-hardware@FreeBSD.ORG>; Sat, 13 Sep 1997 12:13:23 -0700 (PDT)
Received: by gw.rinet.ru id XAA09441;
  (8.6.11/vak/1.9) Sat, 13 Sep 1997 23:03:03 +0400
From: marck@gw.rinet.ru (Dmitry Morozovsky)
Message-Id: <199709131903.XAA09441@gw.rinet.ru>
Subject: Re: Cyrix/IBM 200+ memory speed
To: hlew@www2.shoppersnet.com (Howard Lew)
Date: Sat, 13 Sep 1997 23:03:02 +0400 (MSD)
Cc: freebsd-hardware@FreeBSD.ORG
In-Reply-To: <Pine.BSF.3.91.970913102009.13135B-100000@www2.shoppersnet.com> from "Howard Lew" at Sep 13, 97 10:36:04 am
X-Mailer: ELM [version 2.4 PL24 ME8a]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-freebsd-hardware@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Hi there,

> > here is memory speed determined via dd </dev/zero >/dev/null
> > (CPU still recognized as 486)
> > 
> > Results are very similar with clocks 60x2.5 and 75x2 (BTW, my ASUS T2P4 need
> > to be set up as 60x3 (!) to be 60x2.5 indeed)
> > 
> > Block size	4k	16k	128k	512k	1m	2m
> > dd mem speed	135	113	140	113	50	44
> 
> Hmmm... Shouldn't 75x2 give much better memory speed than with a 60MHz bus?
> Cyrix 6x86MX-PR166 (66x2), PC Chips M560 ALI Alladin IV+
	   ~~~~~
> Block size      4k    16k   32k  64k  128k    512k     1m      2m
> dd mem speed   180    323   282  220   135     117     76      73
> 
> Average values are a little higher if shut down X and use a vty (i.e. 1m
> is 82)

Howard, look twice at underlined statement. My CPU is 8x86/L-PR200+ (not MMX)

However, results are *independent* of bus speed -- and that's strange for me
too. If I use 60x3 (not x2.5! -- then BIOS reported CPU as 150+) results are
the same (differences may be kept inside 1-2%) as for 75x2. Misterious.

Sincerely,
D.Marck
========================================================================
=== D.Marck --- Dmitry Morozovsky --- marck@rinet.ru --- Wild Woozle ===
========================================================================