74 Gig WD Raptor rating

D

Doug

Hey group,

I just bought this drive mentioned in the heading, and I ran "SisSoft
Sandra 2004" and the benchmark only comes back at 18 ms!! Now the transfers
it says are a little over 50 MB/s so I *think* that's pretty good. What has
me puzzled is the average seek time it brings back. Is this drive working
properly and its merely a glitch in Sandra, or is that correct for the file
size sandra is using? I would certainly think it would be better then that
no matter what the test file size was though.

System spec's:
ASUS P4C800 Deluxe
P4 2.8 / 512 (OC'd to 3.08 Ghz) 800 FSB
2X512 matched Dual Channel DDR PC3200
Pioneer DVD 109D DVD Burner
550 Watt true Thermaltake PS

Everything seems to be functioning perfectly and I actually can tell the
hard drive is loading very fast compared to my old system, but as I stated,
the benchmark has me perplexed. I just wanna make sure I get the most out
of my new drive! Would there possibly be something in the BIOS that I
overlooked that could be holding it back, IF its being held back? I set
all the S-ATA per the ASUS book and all seems fine after the 10th check.

Is there another benchmark possibly that gives the seek times in ms's
instead of peak burst rates and the what not?
I'm not a genius at hard drives, but not a rookie either.

TIA,

Doug
 
R

Richard Dower

What is the stripe size?...for speed you need to set the stripe size low.
Are you running two drives in RAID 0?

I am getting this same drive on Friday, so i am kinda interested here
myself.
 
R

Richard Dower

Doug said:
Hey group,

I just bought this drive mentioned in the heading, and I ran "SisSoft
Sandra 2004" and the benchmark only comes back at 18 ms!! Now the transfers
it says are a little over 50 MB/s so I *think* that's pretty good. What has
me puzzled is the average seek time it brings back. Is this drive working
properly and its merely a glitch in Sandra, or is that correct for the file
size sandra is using? I would certainly think it would be better then that
no matter what the test file size was though.

I read somewhere that you have to set some different file size to get an
accurate readout.
 
D

Doug

As far as the stripe size, I have no idea, I just set this thing up per
the ASUS manual and wahhh lahhhh! I only have the single drive so no RAID
*YET*! *evil grin*. I"ll keep checking the posts and hopefully someone
will have some insight into this. BTW: The advertised Average Seek Time
for this drive is 4.5ms....

Thanx again for reading everyone!

Doug
 
R

Richard Dower

Doug said:
As far as the stripe size, I have no idea, I just set this thing up per
the ASUS manual and wahhh lahhhh! I only have the single drive so no RAID
*YET*! *evil grin*. I"ll keep checking the posts and hopefully someone
will have some insight into this. BTW: The advertised Average Seek Time
for this drive is 4.5ms....

Thanx again for reading everyone!

Doug

When doing RAID 0 i have heard a good stripe size is 32K. It depends on what
you are doing with the PC. If you're doing video editing or large file
transfers i have heard a stripe of 128K is best.

I'll try the 32K stripe and see how it moves, i'm wondering...i currently
have an 80GB Seagate SATA drive. Thus i am wondering if i will speed a nice
speed increase with the Raptor?

Did you read any reviews in regards the average seek time?
 
S

Spajky

I just bought this drive mentioned in the heading, and I ran "SisSoft
Sandra 2004" and the benchmark only comes back at 18 ms!! Now the transfers
it says are a little over 50 MB/s so I *think* that's pretty good. What has
me puzzled is the average seek time it brings back. Is this drive working
properly and its merely a glitch in Sandra, or is that correct for the file
size sandra is using? I would certainly think it would be better then that
no matter what the test file size was though.

test your older disk /setup/ against your Raid0 Raptors with same
program but use Disk bench/ no of points you get (index!) in each case
& compare. This way you will see the true life comparision between
setups.
My 20G/7,2k/2Mb Quantum gets 18.000 points /sandra2k1pro/ ;
you should get at least double! (that means once faster than mine!)

-- Regards, SPAJKY ®
& visit my site @ http://www.spajky.vze.com
"Tualatin OC-ed / BX-Slot1 / inaudible setup!"
E-mail AntiSpam: remove ##
 
J

jpsga

Well Doug .. My friend Richard ( like the rest of us ) sees Raptor and
thinks RAID 0.
In your case the Access time does seem a little high but Sandra is a little
week in this area. Access time=seektime+rotational latency+controller delay.
My 36G units run about 12ms with Sandra, HD Tach reports8.9ms.and AIDA32
agrees with HD Tach. Again, we don't *how* any of these people make the
measurement. The 36G drives are in RAID 0 (with a 32k stripe) and get a
sequential read speed of about 100 mB/s.
So, in a non-RAID configuration, it sounds like you are right where you
wanted to be.
jps
 
D

Doug

Okay, hey, I really appreciate your input and maybe I'll see about using
another benchmark and see what that produces. I actually am pretty sure its
working properly cause my bootup time with XP Pro with 34 processes to sort
is under 30 sec's. Compared to over a minute before with my old system. I
would just like to see the seek times in an actual benchmark under 6ms for
peace of mind that the thing is correct.

But once again, thank you for your time!

Doug
 
R

Richard Milser

Hi guys
Am in UK - magazine herepersonal computer world had mini review of hard
disks in their May 04 issue. The WDraptor WD740GD 74GB SATA150 was included.

To summarise :
got reasonably good review
little pricey ; gets little hot; probably best used in RAID arrays.
Average seek time was 5.2ms which was best in test.
It performed best in the 2 benchmarks they used too:

PCmark04 hard disk drive suite overall score 5670
PCmark04 application loading (Mbytes/sec) score 8795

This was compared to a number of other SATA and ATA drives

Hope this is of use to you


Richard
 
D

Doug

Actually Richard, I was more concerned with what benchmark *I* could use to
test *my* specific drive to make sure it was functioning properly. I got a
bad report from SiSoft Sandra 2004 and was concerned I wasn't getting the
full speed I was capable of and was wondering mostly, what I could do to
test its functionality. For *MY* drive of course.

Seems as though a lot of guys have comments about the drive, but nothing to
help really with my specific problem, if any. I have put a hard drive
cooler on it before I even fired it up for the 1st time for I've already
read many many reviews before my purchase, and most of them say that its a
hot running drive, so I took precautions.

As I said, it *seems* to be working perfectly, but I just wanted to confirm
that it is.

Once again, all comments welcome and thanx to everyone for their help thus
far!

Doug
 
R

Ron Reaugh

Doug said:
Okay, hey, I really appreciate your input and maybe I'll see about using
another benchmark and see what that produces. I actually am pretty sure its
working properly cause my bootup time with XP Pro with 34 processes to sort
is under 30 sec's. Compared to over a minute before with my old system. I
would just like to see the seek times in an actual benchmark under 6ms for
peace of mind that the thing is correct.

Benchmarks generally don't measure 'seek time' but 'average access time'
which is different and longer. You'll never see a value near 6 ms. for
these drives.
 
R

Ron Reaugh

Doug said:
Actually Richard, I was more concerned with what benchmark *I* could use to
test *my* specific drive to make sure it was functioning properly. I got a
bad report from SiSoft Sandra 2004

Sandra is useless as a disk benchmark.
 
S

Spajky

Sandra is useless as a disk benchmark.

IMHO you are wrong:

its drive index shows for average user (who does not properly
understand other results (tech gibberish) getting out of benchmarks)
the best average comparision between drives & setups, because it
relies on a complex algorithm while benchmarking (read/write of
samples of different programs) & whole drive subsistem (HD cache, Rpm
speed, firmware, Vcache in OS, DMA, & UDMA mode, density on the
platters, mecchanical latencies etc.) in practical use (life)!

It is NOT pure drive bench, but it cleary shows difference of two HD
setups (PCs). For example, if the first sistem has an index of 20.000
& the other 40.000, means that disk operations with the second setup
is once faster than the previous. [another example: if you disable DMA
you will get only half of the result approx.!] This kind of difference
you tipically get with single 80Gb 7,2k drive & 2 Raptors in Raid0 for
example!

you can simply verify this by copying a large folder with mixed data
back & forth between drives on different IDE chanells ! (if comparing
MB/s speed /divide folder size with average time/, you´re probably get
values under 10Mb/s !!! This CLEARLY SHOWS what is the REAL
BOTTLENECK in modern PCs- meccanical stuff ! (solution for very
rich people: get the biggest solid state (flash) IDE drive! :)

practical HD speed mostly depends on HD Rpm speed & data density on
its platters; mostly all other like UDMA33-133 is mostly marketing
stuff (well, for Raid0 UDMA66 & further is adviceable!); theorethical
max speed of data transfer between HD cache & chipset really does not
impact much (2-3%) on HD performance!

Since a lot of people have Sandra installed & making bench with it, I
wonder why mostly post their MB/s results & not the drive index which
would be best comparision between sistems!

But there are other Complex Benches too & good also for the whole
sistem like PassMark´s Perf.Test etc ...



-- Regards, SPAJKY ®
& visit my site @ http://www.spajky.vze.com
"Tualatin OC-ed / BX-Slot1 / inaudible setup!"
E-mail AntiSpam: remove ##
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top