From owner-freebsd-chat Wed Dec 15 16:58:17 1999 Delivered-To: freebsd-chat@freebsd.org Received: from smtp01.primenet.com (smtp01.primenet.com [206.165.6.131]) by hub.freebsd.org (Postfix) with ESMTP id 6BC49155EB for ; Wed, 15 Dec 1999 16:58:08 -0800 (PST) (envelope-from tlambert@usr09.primenet.com) Received: (from daemon@localhost) by smtp01.primenet.com (8.9.3/8.9.3) id RAA14552; Wed, 15 Dec 1999 17:57:51 -0700 (MST) Received: from usr09.primenet.com(206.165.6.209) via SMTP by smtp01.primenet.com, id smtpdAAAwTaqdC; Wed Dec 15 17:57:36 1999 Received: (from tlambert@localhost) by usr09.primenet.com (8.8.5/8.8.5) id RAA28775; Wed, 15 Dec 1999 17:57:34 -0700 (MST) From: Terry Lambert Message-Id: <199912160057.RAA28775@usr09.primenet.com> Subject: Re: dual 400 -> dual 600 worth it? To: brett@lariat.org (Brett Glass) Date: Thu, 16 Dec 1999 00:57:34 +0000 (GMT) Cc: tlambert@primenet.com, dscheidt@enteract.com, ragnar@sysabend.org, noslenj@swbell.net, chat@FreeBSD.ORG In-Reply-To: <4.2.0.58.19991215173331.046e1aa0@localhost> from "Brett Glass" at Dec 15, 99 05:36:25 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-chat@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > > >The glue needed to build an N-way > > > >machine will always be less expensive than N uniprocessor boxes. > > > > > > Not so. The special chip sets are usually priced at a premium. > > > >I think this is because they work, and allow things like more > >than 2 PCI bus masters at a time, compared to many chipsets, > >whose arbitration logic fails over 2 PCI masters. > > That's correct. Most of these chipsets are produced in relatively > small volumes by server manufacturers, who must devote a lot of > time, effort, equipment, and staff to R&D. One pays a premium > for that! My point here was that I don't give a damn how cheap it is, if it doesn't work. It doesn't matter if I'm getting a palm computer, a pager, or installing a network operations center: if it doesn't work, it's not useful for anything but landfill. > The most cost-effective solution, when one needs more computing > resources than fit cheaply into one box, is to find ways to > distribute the problem cleanly among MANY boxes. SMP is, most > of the time, either a last resort or a way to throw money at > the problem rather than finessing it. I'm not even involved in your SMP thread; I'm only saying that what's "special" about the chipsets you seem to find too expensive is that they actually _work_, compared to the cheaper chipsets you are putatively defending. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message