From owner-freebsd-hardware@FreeBSD.ORG Sat Jun 21 01:51:36 2003 Return-Path: Delivered-To: freebsd-hardware@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id A777437B401; Sat, 21 Jun 2003 01:51:36 -0700 (PDT) Received: from h24-68-151-112.ed.shawcable.net (h24-66-229-2.ed.shawcable.net [24.66.229.2]) by mx1.FreeBSD.org (Postfix) with ESMTP id E954643F75; Sat, 21 Jun 2003 01:51:34 -0700 (PDT) (envelope-from soralx@cydem.org.ua) Received: from localhost (localhost [127.0.0.1])h5L8nE7F017592; Sat, 21 Jun 2003 02:49:14 -0600 (MDT) (envelope-from soralx@cydem.org.ua) From: To: andreas@freebsd.org Date: Sat, 21 Jun 2003 02:49:14 -0600 User-Agent: KMail/1.5 References: <20030615201628.GA2120@titan.klemm.apsfilter.org> <200306160026.44056.soralx@cydem.org.ua> <20030616070541.GG2974@titan.klemm.apsfilter.org> In-Reply-To: <20030616070541.GG2974@titan.klemm.apsfilter.org> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200306210249.14213.soralx@cydem.org.ua> cc: freebsd-hardware@freebsd.org Subject: Re: Which 160-180 GB ATA disk is reliable and fast ? X-BeenThere: freebsd-hardware@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: General discussion of FreeBSD hardware List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 21 Jun 2003 08:51:37 -0000 > Well since I can't get real room for a new ATA controller > I think I'll go with IBMs .... > The dada density should be lower on their 180 GB drives, > they have 6 heads instead of 4 on Seagate.... and they also have tagged queuing and glass platters (which is good, but we don't know how well they were tested) probably, it's the best choice plz report if they fail :) > > > I'll connect the drive to my on-board ATA interface which is > > > only capable of UDMA-66. > > > > This is not good. If you connect UDMA100 HD to UDMA66 interface, the > > performance of the drive decreases signifacantly and non-proportionally > > (I'm not sure exactly why it is so now) > > Well the Seagate UDMA100 on a UDMA 66 BUS was much quicker as > the UDMA 66 disk that came with the machine.... > They both were primary ..... > So I can't second that .... Maybe its even faster when using > a real UDMA 100 controller, but I couldn't notice such a > degradation as you mention. > > Maybe only with certain drives ??? Yes, I remember now that it is a problem with Seagate Barracuda ATA IV. I also recall that it shows very low performance when works in UDMA mode _lower_ than 4 (UDMA/66), so you're still OK. UDMA/100 HD will still be a bit slower on UDMA/66 controlle than on UDMA/100 one. Look, for example, on 7200 RPM drive with 8Mb cache. To empty its cash with UDMA/66 interface it takes at least (7200/60)*(8/66)=14.(54) revolutions, while on UDMA/100 it'll take (7200/60)*(8/100)=9.60 rev's, which is ~5 revolutions less. The performance also depens on how efficient is the drive's caching algorithm. > > > and performance is also a matter. > > depend on your application - most of the modern HDs have minor > > performance differences > I do homerecording under XP and "make worlds" ;-) then IBMs are very good for you 21.06.2003; 02:48:40 [SorAlx] http://cydem.org.ua/