In article <hAAIf.16130$7C3.13392@trnddc08>, Wes Newell says...
http://www.tomshardware.com/2006/02/...ends_performan
ce_lead/page7.html

No they weren't early SATA drives.

Summary:
http://www.tomshardware.com/2006/02/...ends_performan
ce_lead/page14.html


--
Conor,

Same shit, different day.

On Wed, 15 Feb 2006 14:35:03 +0000, Conor wrote:


Well, I don't know what charts you are looking at, but this one clearly
shows the raid0 array kicking the raptors tail in min., aver., and max.
read MBps 129.3 to 86.9. And they didn't even post the raid0 write
results in the second chart. Only single drives with the raptor of course
winning that by not much of 85 to 71 MBps.

http://www.tomshardware.com/2006/02/...ead/page8.html

The next one they show with both in it is the interface performance where
the raid 0 beats the single raptor soundly again 164 to 126. Even other
single sata 300 drives beat the raptor in this since it's still a standard
sata 150 drive.

The raptor did marginally win the sysmark app test, but that's not a good
benchmark for drive performance. And if you based it the winner on that
benchmark you're just fooling yourself. And the cost of the Raptor is
rediculous. I like the way they spun the so called win;

"The Raptor-X's performance is even good enough to beat a RAID 0 array
consisting of two modern 7,200 RPM drives, except in terms of pure
throughput, of course."

The raptor won one appalication benchmark by a small margin and 0 actual
drive throuput benchmarks. The spin doctors at work.:-)

--
Want the ultimate in free OTA SD/HDTV Recorder? http://mythtv.org
http://mysettopbox.tv/knoppmyth.html Usenet alt.video.ptv.mythtv
My server http://wesnewell.no-ip.com/cpu.php

Bill wrote:

I have a system that runs a RAID0 array using two WD160JS. Cost of
these two drives is about the same as a 74gb Raptor. Rest of the system
is a Gigabyte K8N51GMF-9, 2x512 Corsair XMS CAS2, 3700 San Diego o/c'ed
to 2.5ghz (obviously more an office machine than a game machine) I have
very little lag in anything that is i/o dependent.

I can't compare it head-to-head with with Raptor machines, but the
computers I've used with Raptors don't compare.

That said, my next system will likely use three or four drives in RAID0
to take full advantage of even more powerful processors. There may be
some increased risk, but we all need to regularly back up our important
data.

I usually take ANY review with a grain of salt but do compare measured
results. I also compare any reviews from Tom's to those on AnandTech.
I don't often trust manufacturer's cnumber since they just might tend to
be a little biased.

Bill wrote:

Nope. Going for speed. 0+1 only provides the same speed as 0. Four
drives in RAID0. The only point where the speed would be the saem is
the initial access since seek time isn't helped by dividing the load
across multiple drives.


I wouldn't go that far. They might be doing as a lot of other
organizations have been by providing draft reviews to the manufacturers
involved. It can help keep them from forking out more in legal fees and
damages than they could possibly make from advertising.

Under any circumstances, I only look to actual measured results and only
glance the editorial comments to see if there are failures, installation
issues, compatibility problems or if the product falls below expectations.

I guess this comes from 50 years of having to read past manufacturer's
"claims".

Bill wrote:

Nothing special. Office apps, browsing, newsgroups, email, a few
lightweight games. Its just that load times are next to nothing, file
read and write times are just really quick.



Check out the legal actions initiated by Bose if you think that being
honest prevents law suits.




Not that I've been involved very long, but I had an original IBM PC with
a VERY LOW serial number that only could handle single-sided 160k floppy
drives. Bought it new BEFORE the double-sided floppies were available.

Old fart? Yeah, pretty much so. Kids are finishing college in the next
year or so. But if you've ever bought anything on the net, you've
likely used a technique that I either invented or worked on.

Bill wrote:

Rather than doing RAID1+0, you could opt for an inexpensive IDE drive to
use for back-up of your data. Periodically copy all of your data to the
back-up. A lot cheaper and easier to manage. There is an extra
possible benefit of putting the swapfile in a small first partition of
the new drive. That space has faster access and read/write speed. The
rest of the drive won't be accessed very much so it won't interfere with
other access.

Since reand and write speeds are a lot slower than the interface speed,
the fact that the transfer is across the IDE channel won't slow the system.


They are THE WORST!!!



Nope. Not all my fault. I only worked on a few methods of securing
e-commerce. I've been appropriately punished for that.

"Bill" <bill@c.a> wrote

In my experience 99% of problems that necessitate an image restore
have nothing to do with drive failure. It's stuff like botched software
installs and updates, driver installs and updates, viruses etc, none of
which are recoverable with RAID1. The corruption or whatever
simply gets copied to the mirror.

It's much better to use a second set of drives for a periodic image
backup, using Ghost or other imaging software.

Bill wrote:

Sounds like you could build a USB drive using an inexpensive IDE drive.
With the speed of USB, you wouldn't have to worry about a fast drive
or big cache. Saves a few dollars.