From owner-freebsd-hackers  Fri Apr 11 17:37:20 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id RAA03658
          for hackers-outgoing; Fri, 11 Apr 1997 17:37:20 -0700 (PDT)
Received: from caliban.dihelix.com (caliban.dihelix.com [198.180.136.138])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA03648
          for <hackers@freebsd.org>; Fri, 11 Apr 1997 17:37:17 -0700 (PDT)
Received: (from langfod@localhost) by caliban.dihelix.com (8.8.5/8.8.3) id OAA01374; Fri, 11 Apr 1997 14:35:05 -1000 (HST)
Message-Id: <199704120035.OAA01374@caliban.dihelix.com>
Subject: Re: 430TX ?
In-Reply-To: <Pine.BSF.3.95.970411152235.27533A-100000@harlie.bfd.com> from "Eric J. Schwertfeger" at "Apr 11, 97 03:24:22 pm"
To: ejs@bfd.com (Eric J. Schwertfeger)
Date: Fri, 11 Apr 1997 14:35:05 -1000 (HST)
Cc: langfod@dihelix.com, hasty@rah.star-gate.com, steve@visint.co.uk,
        louie@TransSys.COM, michaelh@cet.co.jp, avalon@coombs.anu.edu.au,
        terry@lambert.org, hackers@freebsd.org
From: "David Langford" <langfod@dihelix.com>
X-blank-line: This space intentionaly left blank.
X-Mailer: ELM [version 2.4ME+ PL31 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Eric J. Schwertfeger
>
>
>> (That and why mohterboard makers dont put caches on Pentium Pro motherboards
>>  to interface the slow main memory and the faster on chip cache.)
>
>Can't imagine that a L3 cache would be much use unless it's SEVERAL Meg
>in size.  Since the PPro has at least a 256K L2 cache, the L3 cache would
>hurt unless it's quite a bit larger than this.

Right. If you look at the Mips, Alpha, PowerPC, Sparc, etc.. They all
have 1-2Meg caches on the higher end systems.

Granted there are some major differences in architecture differences
but those differences are beginning to get fewer these days as the
RISC chips get more CISC and the CISC chips get more RISC and
harvard architecure is being used more...... Anyway :)

Main point is that I think even motherboard makers could be more helpful
in improving memory usage. I would pay a premium for a high bandwidth 
motherboard, but I dont see a point in spending $750 for motherboards that
perform just like the $200 boards...

err [RANT off] :)

-David Langford
 langfod@dihelix.com