Date: Tue, 22 Oct 2002 16:15:01 +0200 From: "Harry Kroonen" <h.kroonen@brinktech.nl> To: FreeBSD Laptoppers <freebsd-mobile@FreeBSD.ORG> Cc: FreeBSD Laptoppers <freebsd-mobile@FreeBSD.ORG> Subject: Re: hard drive Message-ID: <3DB57985.5105.6BB988A@localhost> In-Reply-To: <20021021052818.GA554@laptop.6bone.nl> References: <E183LiT-000LYK-00@rip.psg.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Date sent: Mon, 21 Oct 2002 07:28:19 +0200 From: Mark Santcroos <marks@ripe.net> To: Randy Bush <randy@psg.com> Copies to: FreeBSD Laptoppers <freebsd- mobile@FreeBSD.ORG> Subject: Re: hard drive > On Sun, Oct 20, 2002 at 12:26:25PM -0700, Randy Bush wrote: > > for me, the bottleneck on laptops is the disk throughput. the hard > > drives themselves are slow, and the bandwidth to them sucks. for > > big makes etc., a 10GHz cpu would not improve throughput as much as > > doubling the disk goodput. > > > > is anybody making laptoys with real disk i/o subsystems? > > Imho the gap between desktop and laptop is getting smaller. There are > quite some laptops shipping with UDMA100 now. > > "dd if=/dev/ad0 of=/dev/null bs=4096" gives me 22 MB/s. I don't have a > recent desktop so I don't know how this compares to recent desktop > performance but I don't think it is bad. As some general info: I just did the same (count=10000) on a 'recent desktop' AMD1333 with ATA100 / 7200rpm Maxtor 740-DX drives, ad0 (40GB) giving me 22MB/s, ad4 (80GB) at 34MB/s. Laptop Celeron 500 / IBM HD : 16 MB/s More important in the laptop-arena is the difference in available HD's. With two laptops and two HD's I measured the time for "ls -lR /usr/ports" (and yes, both directories were identical ;-) ): Celeron 500 / IBM DJSA-220 : 1m56s Celeron 1100 / IBM DJSA-220 : 1m37s Celeron 500 / toshiba MK2018 : 6m21s Celeron 1100 / toshiba MK2018 : 5m35s (ls -lR /usr/ports on my desktop -> 0m40s) The toshiba was standard with the 1100-laptop (low-budget, from 2 months ago so I should expect that sort of performance), the IBM was a replacement drive (for the 500) bought separately about a year ago. So it looks to be more a difference of access time than of basic throughput. Anyway, just a few numbers, nothing scientific, Harry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3DB57985.5105.6BB988A>