|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
Subject: Re: hmmmm....
From: Lyndon Nerenberg (lyndon
messagingdirect.com)Date: Tue Sep 19 2000 - 18:21:08 CDT
- Next message: Yanek Korff: "ALIAS disaster"
- Previous message: sthaug
nethelp.no: "Re: hmmmm...."
- In reply to: Liviu Daia: "Re: hmmmm...."
- Next in thread: Brad Knowles: "Re: hmmmm...."
- Next in thread: Harry: "Re: hmmmm...."
- Next in thread: Brad Knowles: "Re: hmmmm...."
- Reply: Lyndon Nerenberg: "Re: hmmmm...."
- Reply: Brad Knowles: "Re: hmmmm...."
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Wed, 20 Sep 2000, Liviu Daia wrote:
> years ago. But the real problem with these disks is elsewhere: the
> controller would still hog the CPU for the full duration of a command.
Not for DMA based controllers (UDMA).
The real problem with IDE is that you can only issue one command at a time
to the disk. Contrast this with SCSIs overlapped commands (tagged
commands) and it's easy to see why SCSI is so much faster.
Bleeding edge ATAPI disks now support tagged commands ala SCSI. If your
kernel supports this ATAPI can be as fast as SCSI, and at a fraction of
the price.
--lyndon
- Next message: Yanek Korff: "ALIAS disaster"
- Previous message: sthaug
nethelp.no: "Re: hmmmm...."
- In reply to: Liviu Daia: "Re: hmmmm...."
- Next in thread: Brad Knowles: "Re: hmmmm...."
- Next in thread: Harry: "Re: hmmmm...."
- Next in thread: Brad Knowles: "Re: hmmmm...."
- Reply: Lyndon Nerenberg: "Re: hmmmm...."
- Reply: Brad Knowles: "Re: hmmmm...."
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]