From owner-freebsd-smp@FreeBSD.ORG Wed Nov 5 11:20:19 2003 Return-Path: Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 92D0716A4CE for ; Wed, 5 Nov 2003 11:20:19 -0800 (PST) Received: from main.gmane.org (main.gmane.org [80.91.224.249]) by mx1.FreeBSD.org (Postfix) with ESMTP id 690BE43FA3 for ; Wed, 5 Nov 2003 11:20:18 -0800 (PST) (envelope-from freebsd-smp@m.gmane.org) Received: from list by main.gmane.org with local (Exim 3.35 #1 (Debian)) id 1AHTCT-0004zV-00 for ; Wed, 05 Nov 2003 20:20:17 +0100 X-Injected-Via-Gmane: http://gmane.org/ To: freebsd-smp@freebsd.org Received: from sea.gmane.org ([80.91.224.252]) by main.gmane.org with esmtp (Exim 3.35 #1 (Debian)) id 1AHTCQ-0004zN-00 for ; Wed, 05 Nov 2003 20:20:14 +0100 Received: from news by sea.gmane.org with local (Exim 3.35 #1 (Debian)) id 1AHTCQ-0007zI-00 for ; Wed, 05 Nov 2003 20:20:14 +0100 From: Jesse Guardiani Date: Wed, 05 Nov 2003 14:20:10 -0500 Organization: WingNET Lines: 72 Message-ID: References: <20031105174438.GA8014@SDF.LONESTAR.ORG> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit X-Complaints-To: usenet@sea.gmane.org User-Agent: KNode/0.7.2 X-Mail-Copies-To: never Sender: news Subject: Re: SMPng question X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: jesse@wingnet.net List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 05 Nov 2003 19:20:19 -0000 Kevin A. Pieckiel wrote: [...] >> Could I buy one of those cheap $300 quad Xeon 500mhz >> compaq boxes on ebay and use it as a 2 Ghz compiler box? > > No, not really; you're going to lose some performance in just > keeping up with multiple processors. In other words, it's not > a linear increase in performance when you add processors. OK. That's what I thought. Do you have a figure on the amount of performance lost managing multiple CPU's? I'm looking at this from a cost perspective, essentially. Take These two computers for example: 4 CPU 500Mhz Xeon w/1M cache and 1Gb RAM for $350: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3054734970&category=1484 2.4 Ghz P4 Intel 800FSB (cache size unknown) and 1 GB Ram for $600: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2763877081&category=51139 --------- Ratio Math for Cost ---------- $350 x ------ * --- $600 100 $35,000 = $600x 58.33% = x ---------- End Ratio Math ------------- So, the 2.4Ghz machine is 58.33% more expensive than the quad 500Mhz machine. If I could get 60%-70% of the performance of the 2.4Ghz machine out of quad 500Mhz machine then I'd consider purchasing the quad 500Mhz machine to be a good deal, considering cost. However: --------- Ratio Math for CPU ---------- 2,400 Mhz 100 ------ * --- 2,000 Mhz (4x500Mhz) x 200,000 = 2400x 83.33 = x ---------- End Ratio Math ------------- As shown by the above ratio math, the quad 500Mhz machine is only 83.33% as powerful as the 2.4 Ghz UP machine, so the quad 500Mhz machine running SMPng would have to only incur 23.33%-13.33% operating losses in order to reach my 60%-70% efficiency goal. :) Anyone have an idea what SMPng per processor loss percentages are? The 2.4Ghz machine will probably blow away the quad 500Mhz machine regardless since it has 400Mhz DDR RAM and an 800Mhz FSB. Still, it would be really neat to see some benchmarks on things like this for future decision making. -- Jesse Guardiani, Systems Administrator WingNET Internet Services, P.O. Box 2605 // Cleveland, TN 37320-2605 423-559-LINK (v) 423-559-5145 (f) http://www.wingnet.net