From owner-freebsd-hardware Wed Mar 4 17:58:16 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA23275 for freebsd-hardware-outgoing; Wed, 4 Mar 1998 17:58:16 -0800 (PST) (envelope-from owner-freebsd-hardware@FreeBSD.ORG) Received: from allegro.lemis.com (allegro.lemis.com [192.109.197.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA23221 for ; Wed, 4 Mar 1998 17:58:00 -0800 (PST) (envelope-from grog@lemis.com) Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by allegro.lemis.com (8.8.7/8.8.5) with ESMTP id MAA11913; Thu, 5 Mar 1998 12:27:46 +1030 (CST) Received: (from grog@localhost) by freebie.lemis.com (8.8.8/8.8.7) id MAA00427; Thu, 5 Mar 1998 12:27:49 +1030 (CST) (envelope-from grog) Message-ID: <19980305122749.17883@freebie.lemis.com> Date: Thu, 5 Mar 1998 12:27:49 +1030 From: Greg Lehey To: Mike Smith , Joao Carlos Mendes Luis Cc: hardware@FreeBSD.ORG Subject: Re: TX Chipset and more than 64M Ram References: <199803042012.RAA09257@gaia.coppe.ufrj.br> <199803042038.MAA22209@dingo.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <199803042038.MAA22209@dingo.cdrom.com>; from Mike Smith on Wed, Mar 04, 1998 at 12:38:26PM -0800 WWW-Home-Page: http://www.lemis.com/~grog Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 4 March 1998 at 12:38:26 -0800, Mike Smith wrote: >> Hi, >> >> What kind of problem could I expect from FreeBSD if I run a TX >> chipset motherboard with 128M RAM ? > > Your performance will suck. Your performance will drop instead of increasing. >> This chipset can only cache >> 64M. Anything other than performance ? Is it possible to force >> FreeBSD to use the low 64M preferentially ? > > Take the top 64M out. > > Seriously, it's going to cost you less to replace the board with one > wearing an HX chipset than the time that the TX board will waste you. The German magazine c't, which I personally greatly respect, did a test of a number of motherboards which can cache more than 64 MB in their issue 4/98. The chipsets tested were: Intel 430TX Ali Aladdin IV+ VIA Apollo VP2 SiS 5582 Intel 430HX VIA Apollo VPX The order is the order of speed in c't's BAPCo benchmark (which, unfortunately, is stronly Microsoft-oriented) with 64 MB main memory (430TX is the fastest, with a rating of 225, compared to 221 for teh Aladdin and the VP2). With increasing memory, the TX performance drops, while the performance of the other chip sets increases. At 72 MB, the TX drops below Aladdin and VP2, at 96 MB (!) below the HX, and by 128 MB, it's down to 204, compared to 227 for the Aladdin (which by this time has left the VP2 behind). All reports say that the cache limit is particularly hard on Microsoft due to its brain-damaged memory allocation; I can't verify this, but I'm prepared to believe it. That would mean that the drop under FreeBSD would be less. I'm currently running a TX board with 96 MB, and while I'm trying to replace it, I can't say that "my performance sucks". Greg To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message