Is there any reason to go to the trouble to make sure my 750 Samsung
and seagate Hds are enabled for for Sata II speeds?

From what i've read so far you may have problems with some
motherboards and only a RAID would get anywhere NEAR needing that
speed. (of sata II).

My Hds are only listed as moving data at about 100 MB/second.

I do move and encode Gigs of video.

I just built my Quad 9450 and it process data at 49+ meg/second.
with 78% cpu use. (dvd shrink-analyze option)

My old AMD 4200 X2 @ 2600Mhz only does 18 meg/second
and 100% cpu use.

Looks like I'm not cpu bound anymore but
my new Hds should be able to move data faster then 49meg/second ???

On a straight Hd to Hd transfer a Gig will take maybe 12 seconds.


Thanks,
jmc

jmc@nospam wrote:
There are a couple benchmarks here. They test sequential access across
the disk surface. A disk is faster at the beginning, than at the end.
The benchmark should give a curve. These should be read-only tests.
(The purchased versions may offer read and write tests.)

http://www.simplisoftware.com/Public...request=HdTach
http://www.hdtune.com/download.html (right column - select the read-only version)

If your software does a lot of seeks (lots of head movement of any
significant distance), that will impact the average transfer rate
achieved.

The first one of these, sold on Newegg on May 28. This is the latest
10K RPM SATA drive.

300GB and $1 per GB.
http://www.newegg.com/Product/Produc...82E16822136260

Transfer rate 123MB/sec down to 74MB/sec at the end of the disk.

http://www.anandtech.com/storage/showdoc.aspx?i=3303
http://www.legitreviews.com/article/699/4/

So benchmark the existing drive, and see where it sits in the scheme
of things.

While a 150MB/sec cable transfer rate sounds adequate, there is some
overhead in the packet format. So the actual achieved max transfer
rate, will be less than 150MB/sec. I don't know if that would prevent all
of the VelociRaptor performance, near the beginning of the disk, from
being used or not.

That drive is kinda small, and there are plenty of other ways (RAID) to
construct a higher performance solution.

Paul

<jmc@nospam> wrote in message
news:tja14495gh1q644bbk12ddkged4ikh208s@4ax.com

Why not do it and see what shakes?

RAID reduces or leaves unchanged the demands on the individual drive for
performance.


Probably under highly idealized conditions.

You are less CPU-bound.

That's about 82 megabytes per second. I wouldn't complain about that! ;-)

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:HL6dndwtUfXpT97VnZ2dnUVZ_jqdnZ2d@comcast.com. ..
It would certainly be true of RAID 0, but then
RAID 0 doesn't really fit the acronym very well does it?
I have been useing RAID 5 and I think that does increase demand.
Everything has gotten so cheap that I might move to RAID 1 next.

On Sun, 01 Jun 2008 22:33:08 -0400, Paul <nospam@needed.com> wrote:

SNIP
Downloading as we speak.

74 to 123MB/sec change, wow.
Hmm, that probably explains whys I've gotten quite different
transfer times on my 1 gig copy tests...
different place on Hd and maybe a litte fragmentation.

I want them to come out with SMALL high speed
solid state disks. 4-8 gig raid 0 for my OS drive!
(Might be able to afford couple of thoses!)

Thanks lot!
jmc

<jmc@nospam> wrote in message
news:a54g445jqvpr4nf3f21668ctfsg9ijmgjt@4ax.com

They are called flash drives. Trouble is that common flash memory tops out
around 6 megs/sec DTR, which is about 1/10 the speed of a modern hard drive.

Arny Krueger wrote:
The drives here look pretty good. Flash performance drops
when you're writing a bunch of small files, but that
isn't being tested here.

http://www.anandtech.com/weblog/showpost.aspx?i=326

Paul

"Paul" <nospam@needed.com> wrote in message
news:g2b5sd$iso$1@aioe.org

Didn't someone say "affordable"?

http://www.neostore.com/detail.asp?ProductID=1192

Arny Krueger wrote:
My purpose in posting that, was to show that flash
doesn't have to be slow. There is a difference between
SLC and MLC flash, and the products benchmarked there
could be SLC based.

The thing I'm curious about, is how the flash industry
expects to crush ordinary hard drives, from a pricing
perspective. Something better than what they're currently
offering is needed, and I've never read any details
explaining how it'll get better.

Paul

"Paul" <nospam@needed.com> wrote in message
news:g2c3nl$3pf$1@aioe.org
Or, they could have simply used a number of flash chips in parallel.

The usual rule of thumb is that the cost of RAM falls by 50% every two years
or so. Additional cost cutting comes when a lower-volume product goes
high-volume and there are economies of scale.

It seems like flash is popular enough for the economies of scale to be
already at hand.

The other factor is the possibility of a public "so what" attitude about
further increases in the size of hard drives.

Other than editing large videos, I don't see a lot of benefits to hard
drives > 1 TB.

On 6/06/2008, Arny Krueger posted this:
These are affordable :-)

Page 65 of the July 2008 issue of Mac|Life magazine shows the same
drives (the 64GB version) at $1959 a pop, compared to $1179 on the
company's web site that you linked to. The 64's have product ID 1193.

I'm still waiting for the price to come down at least an order of
magnitude. I'm continuing to breathe, however :-)

--
Gene E. Bloch (Gino) letters617blochg3251
(replace the numbers by "at" and "dotcom")