A while back I had posted comparing my WD IDE to my SG SATAII. Now I have a RAID 0 with the 2 Barracudas, and thought I'd post some numbers in case anyone was wondering what you could expect. I have found plenty of stuff on raptors, since most of the people that would go to the trouble to post stats probably have a boss rig with the wicked awesome 10k's.. if anyone is wondering how the mainstream stuff compares...

one ST32506AS..






2 ST32506AS in RAID 0 (128K Striping)





not the jump I was looking for, but maybe I will be able to notice more in the long run..

Notice that my sequential read is lower, but sustained read is higher.. Anyone got and info on how that works?

thanks

To be perfectly honest, I've never understood the scores from HDTach for raid arrays. I've run several bench suites on mine and the results from HDTach always leave me scratching my head in that they are not consistent with the other packages. If I have time over the weekend, I'll drop another SATA drive in my case (#2 below) and -*test*-('") it against the four drive array with Everest and see what the results yield.

I run 2 x WD Raptors (2x36.7Gb 16Mb cache) in RAID0 here's everest results:-

Here's the HD Tach result on RAID array:-

And just one WD SATAII 250Gb 16Mb cache:-

How does that compare.

I guess I feel a little better.. I expected these drives to come in somewhat slower than raptors, but I didn't have a good feel for what that would be. The thing that got me was that the burst speed was lower on my array, but I guess that parameter doesn't really mean much for real world performance.. thanks for the post!

Update..

I followed WGF's tweak here:
http://asustech.15.forumer.com/index.php?s...845&hl=deskstar

and here's what it did for me..

before (read caching enabled)


after (no caching)



so.. does that mean you're wasting your money buying a 16mb cache HD, or is this just a windows/software caching kind of thing?

Thanks again!

EDIT: I just read the rest of the other post, and I'm re-enabling the cache so hopefully I don't corrupt my array.. Thoughts?

I don't think you risk corrupting your array by turning off the read caching. Don't ever shut off write caching, or you'll cook it for sure.

There is no doubt in my mind that the 16mb cache disks are faster than the 8mb cache disks. I have six Raptors of the former variety and four of the latter. The newer ones are faster, hands down.

When you shut down read caching, you might gain something in the world of benchmarks, but you don't gain anything in the real world performance of your disk(s). In the real world, you're not reading from start to finish on your disk(s). You hop and skip around. To the extent that your disk(s) can cache data on a read ahead of when the processor needs it, you're ahead of the game.

When the masters speak, I listen... thanks..

I'm planing on a RAID 0 using 2 WD1600YS drives.
Will post some feedback on it.

Just ordered the WD1600YS.