ATA100 or ATA133 RAID 0?

[KSChan]

I am using a single harddisk Western Digital WD800JB
which is nearly full all the time~
>always less than 10GB free space
>>sometimes less tha 2GB free space

My Mainboard is Gigabyte GA-8IHXP(3.0) which has
"Promise IDE RAID 0,1 and Ultra ATA 133 support"

My computer's configuration:
P42.53G(533FSB)
512MB PC800 16bit Rambus

NOW,
I want to buy two 120GB harddisk(8M Cache)running in RAID 0.

There are several options:
(1) Western Digital WD1200JB
(2) Maxtor 6Y120P0 (*UDMA133)
(3) IBM IC35L120AVV207-1
(4) Hitachi 722512VLAT80

For a single harddisk,
which option has the best performance and stability?

Is UDMA133 faster than UDMA100 in RAID 0?
If so, would option(2) better than option(1)?

Overall,which option is the best for running RAID 0?

 

[smitbret]

I love my Western Digital 1200JB Drives in RAID 0. The ATA 133 will offer no advantage over the ATA 100. You'll probably do very well with whichever is the least expensive right now.
-Brett

 

[siranthony]

I can't say for sure that ATA133 is better than ATA100. But I did have a promise ATA100 controller. My new MB came with a VIA ATA133 controller. I can say my new ATA133 controller has much better bench marks. In fact it beats SATA 150 RAID acording to Sandra.

 

[svol]

No there is no single performance difference... it is all marketing hype. A HD will never fully utilize ATA100 speed (it will go barely over ATA66 speed) so ATA133 is just total BS. And besides ATA133 isn't even an official standard IIRC.

So I would get the Western Digital drive which also has 3 years warrenty instead of the 1 year the other companies offer.

I love my Delta 60HP 7000 RPM fan that puts out more dB then CFM

 

[Cloaked]

that sandra comparison thing is bs. It says that my 40 gig 7200 rpm ata100 maxtor ntfs drive is faster than 15000 rpm scsi drives in raid 0

 

[siranthony]

Hmm, Far out. What benchmark (free) would you recommend. I'll post the results.

 

[Cloaked]

im not sure, many sites use iotach and some other benckmarks, but you might have to buy them. Your best bet is just looking at loading times... hmmm

 

[smitbret]

Win Bench '99 seems to give me the most accurate and consistent results. It also works well on RAID setups. Many benchamrk programs don't.
-Brett

 

[siranthony]

Ok,
Here are the results from winbench99. I'm only posting a couple as there are like 10 scores. I have no idea how these relate to other systems. Please let me know how they stack up.

Buisiness disk mark
before diskeeper after

10500 13900

Highend disk mark
before diskeeper after

37500 48500

 

[smitbret]

In a RAID 0 setup with two ATA 100 drives, would the theoretical max rate be 200mb/s, or would it still be 100mb/s. After reading this thread, I benchmarked my RAID 0 setup for a few minutes yesterday. I did notice that the start of the graph was almost exactly at 100mb/s. Of course it tapered off as te BM ran. Could that have been the ATA 100 ceiling that I hit, or just the physical limits of the drives?
-Brett

 

[svol]

Hmmm... interesting question. You got two drives with ATA100 transfer rate... if they're at different cables then you could get 200 MB/s transfer rate. But ofcourse this will never be reached because:
1. You will hit the physical limit of the drives much faster.
2. The PCI bus only can transfer 133 MB/s in total (so if you play sound at the same time it is actually a little lower).

I love my Delta 60HP 7000 RPM fan that puts out more dB then CFM

 

[smitbret]

Aha, so it was the physical limit. That's what I figured but I wanted a second opinion before I accepted my own reasoning as truth.
-Brett