Real / measurable performance gain for RAID 0 array

[ethu]

I want to know as to how much of real / measurable performance gain can be expected by having OS/applications on stand-alone HDD versus having it on two HDD in hardware based RAID 0 array.

(all other things being same in the machine).

The machine configuration is as follows.

Motherboard - ASUS P5K PREMIUM/WIFI-AP LGA 775 Intel P35 ATX Intel
Memory - 8 GB (Two kits of Patriot Extreme Performance 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model PDC24G6400LLK)
OS - Windows 7 64 bit RC

 

[r_manic]

Have you tried consulting any articles under the Storage section of the main Tom's Hardware site?

 

[daship]

The Raid 0 will pown it all day and all night.

 

[blackened144]

The Raid 0 will pown it all day and all night.

That depends on your definition of pown. Back in the day I had 2x80 in Raid0 and didnt notice any difference at all. This was back in the day though and both drives were running on the sane IDE channel which hindered performence. But i did it again recently with 2x400gb SATAII drives and I could barely notice the difference in everyday operations... I recently moved to raid0 SSD drives though and those KILL any other drive solution I have ever had.. I havent tried running on just 1 SSD to know how much difference the raid0 makes though.

EDIT: I meant to add that raid0 does not really shine when used for the os with traditional disks. It really shines in large sequential file transfers, which the os rarely does. I use my 2x400 and 2x500 gb arrays to handle extremely large archives. Extracting from one array to the other shows almost a linear doubling of speed compared to a single disk to single disk extraction and in between 3-4 times faster than extracting from one single disk to the same single disk.