- Promise Ultra100 vs Promise Ultra133
- Posted by _R on July 18th, 2005
Has anyone noticed any compelling reasons to use one or the other of
these cards? Subjective opinions on whether the 133Mhz is a benefit
over 100Mhz? Problems with either?
- Posted by Folkert Rienstra on July 18th, 2005
"_R" <_R@nomail.org> wrote in message news:6l2nd11j5hf4a9osd531vog68ej35iuqnj@4ax.com
Not for a single drive per channel.
Not for 2 drives per channel either unless you use them in tandem:
multiple drive searches, Raid etc.
- Posted by Rod Speed on July 18th, 2005
_R <_R@nomail.org> wrote
Pass.
It isnt with hard drives themselves, so wont be with the cards.
Pass.
- Posted by Tod on July 18th, 2005
In real world performance
An ATA/133 hard drive would not have any faster transfer speed over a
ATA/100 hard drive
The Promise ATA/133 card could just be a later revision of the Promise
ATA/100 card.
Any bugs that showed up in the ATA/100 card would have a chance of be fixed
in the ATA/133 card.
"_R" <_R@nomail.org> wrote in message
news:6l2nd11j5hf4a9osd531vog68ej35iuqnj@4ax.com...
- Posted by Arno Wagner on July 19th, 2005
Previously _R <_R@nomail.org> wrote:
The 133 version is a bit fatser when running two drives per channel in
a Linux software RAID. But since this slows down the drives
significantly, it is far better to have only one drive per channel. No
noticable difference between the two controllers with one drive per
channel.
Arno
- Posted by _R on July 19th, 2005
On Mon, 18 Jul 2005 16:46:49 -0700, "Tod" <no_spam_me@comcast.net>
wrote:
Thanks to alll for your replies.
I wouldn't have thought there would be much advantage aside from
getting a quick 8MB or so from drive cache. And I think benchmarks
have shown that even that relatively large on-drive cache has limited
value.
Re firmware: Both the 133 card and the 100 are flashable. I think
they are about even as far as BIOS and driver rev goes. The 100 has
an Eprom and a 33Mhz (or 66mhz?) crystal that's not present on the
133, so I suspect that one motivation was trimming the cost. The 133
still has PC traces for the xtal, etc but no parts soldered in.
I've also heard of people converting Ultra100 to a FastTrak 100 RAID
by changing a resistor or two. I have both boards, and again they do
appear identical, even to components. I have no use for software
raid, so I was actually considering a reverse change.
(Motivation: I need to get several systems running with the same ATA
controller, and no one seems to make a good up-to-date straight ATA
controller. The Promise series is 'OK' and better than newer Silicon
Image boards I've found).
Rumor has it that the FastTrak 100 will work as an Ultra100 (straight
ATA) out of the box, but of course Promise tech support says this
won't work, and that drive signatures are needed even in JBOD.
- Posted by Folkert Rienstra on July 19th, 2005
"Tod" <no_spam_me@comcast.net> wrote in message news:wsGdnUEUCMD2okHfRVn-1Q@comcast.com
In theory it does, just because the interface is chosen dependent
on the STR that the drive can master.
An ATA-100 HD would have an STR of less than ~45MB/s ((100/2)-10%).
An ATA-133 HD an STR of ~45MB/s up to 60MB/s ((133/2)-10%).
But, in practice the HD manufacturer may have chosen not to go_with/
invest_in the ATA-133 interface when the next drives will likely be SATA.
Well, they all are more or less optimized versions of previous ones.
That's funny. You would have thought that the latest version of
the IDE card would by now be faultless if that were the case.
Obviously new bugs emerge with newer cards and the ATA-133
card may have a bug that the ATA-100 card doesn't have.
