Re: tlug: SCSI drives vs. ATA

[Dennis McMurchy <denismcm@example.com>]

On Sun, 6 Jun 1999, Jonathan Q wrote:

> On Sat, 5 Jun 1999, Dennis McMurchy wrote:
> 
> >  Sorry not to reply to the message itself, but it has been bothering
> >me for days that Jonathan Q claimed that SCSI drives were no match
> >for ATA.
> 
> I said what?!?!
> 
> Allow me to re-post:
> -----
> That solely depends on whether or not the SCSI drive is faster than the
> IDE drive.  An old, slow SCSI drive will get its doors blown off by a
> fast, new IDE drive.  
> 
> These days, SCSI and IDE drives are generally mechanically the same,
> with only the drive electronics being different.  On a single drive
> system, you'd be hard put to notice a performance difference.  On a
> dual-drive system, you can put one IDE drive on each header and probably
> see superior performance to two SCSI drives attached to the same SCSI
> adapter.  If there are two IDE drives on the same header, UW SCSI might
> have somewhat of an edge, but I'm not sure.  For drive performance, I
> see little reason to use SCSI in a desktop machine.  The nice things
> about SCSI are that you can hook a lot of devices up to one adapter, and
> they aren't limited to only disks, so you can put a scanner and tape
> drive on there as well.
> -----

  Sorry, Jonathan, your position was clearly much more nuanced than I
had remembered. I was completely mistaken when I characterized it the
way that I did. Your re-posting sets the record straight on that.

  But this still leaves unanswered my question about the performance of
the newest 10,000 rpm screamers, which were only available in SCSI
versions when I last looked a few months ago. These _may_ be an
exception to your undoubtedly correct statement that "these days, SCSI
and IDE drives are generally mechanically the same...".  I'm just
curious.  My guess would be that one of these drives on an U2W/UW bus
would be a good bit faster than any ATA drives available.   

  Once again, Jonathan, I'm sorry for attributing to you statements
that you clearly did not make.  

  Cheers,

Dennis McMurchy, 
Sointula, B.C. / Tojinmachi, Fukuoka
Canada           Japan



-------------------------------------------------------------------
Next Technical Meeting: June 19 (Sat), 18:30  place: Temple Univ.
*** Topic: Linux SMP/Quad Xeon Server
Next Nomikai: July 16 (Fri), 19:30 Tengu TokyoEkiMae 03-3275-3691
-------------------------------------------------------------------
more info: http://www.tlug.gr.jp        Sponsor: Global Online Japan