Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 14 Jan 2004 06:24:37 -0600 (CST)
From:      James Van Artsdalen <james@jrv.org>
To:        freebsd-amd64@freebsd.org
Subject:   ATA (Re: AMD64 support in 5.x versions)
Message-ID:  <200401141224.i0ECObwr093993@bigtex.jrv.org>
References:  <Pine.GSO.4.44.0401071115150.8910-100000@odin.cs.kun.nl> <17426.6989030459$1073697945@news.gmane.org> <btuv0n$2hb$1@sea.gmane.org> <20040113070204.GA764@curacao.n2it.nl>

next in thread | previous in thread | raw e-mail | index | archive | help
> Date: Tue, 13 Jan 2004 08:02:04 +0100
> From: Bill Squire <billsf@curacao.n2it.nl>

> I've got several large SCSI disks spinning and have since purchased an 
> ordinary ATA-133 IDE and a SATA drive.

SCSI vs. ATA can be a religious argument.

On the practical side if you have more than 4 GB of RAM I don't know
of any cost-effective solution that can address memory above 4 GB
directly, without using bounce-buffers.  I understand that some
Adaptec SCSI host adapters can do this (PCI "dual address cycle" or a
full 64-bit bus if available) without resorting to exotic RAID
controllers.

Highpoint sells an RR-1820 PCI-X (64-bit 133 MHz) serial ATA controller
but I cannot get them to send me any technical documentation and
cannot get it to work with Windows with the supplied drivers.  Silicon
Image has a 3124 chip which is PCI-X and ought to be able to directly
address more than 4 GB of RAM but I don't know of any card that uses
this yet.

I have 8 GB of RAM and use a HighPoint 1542 sATA controller, and am no
doubt exercising the bounce buffers extensively.  This seems entirely
stable but for performance I wouldn't want to do it with a disk-bound
system in a production environment.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200401141224.i0ECObwr093993>