SSD Raid and other comments

Question

I spent the time to test a few situations for myself because there is no real life data that I can find on the internet which pertains to a real Mac user environment. These are only my finding and observations so I can finally stop ripping into my 2008 Mac Pro and changing things around constantly. Its not very scientific but to me I don't mind since the Mac user environment is judged upon feel as a large consideration. Also I'm not a high end power user to work with expensive software solutions. Just an average Mac user since '86, interested in speed and feel for speed.
First: I wanted to utilize my 6 sata ports for hard drives because one, I test and need the extra hard drives and two, I was thinking of Raid 0 with SSD. So I researched and found nothing about using a sata optical connected to the ide controller in top compartment of the Mac Pro. So I decided to try anyway and purchased a Manhattan Sata 300 to ide converter. The only reason I bought this item is their web site said Windows, Mac and Linux compatible. I set J3 to cable select and connected the provided y-type power connector to the tiny controller card and the optical. I tried an LG CH08LS10 BD reader/DVD burner and a Samsung SH-B083. Without going into lots of detail everything I threw at the opticals worked. ASP shows ata connected device and of course nothing under sata.
Second: I have read all there is to read about raid 0 and SSD for the boot drive.I have also tried every combination of SSD raid 0 and concluded only based upon observation and feel concerning speed that it made no sense to raid 0 SSD and lose a good sata port. Yes it does have an improvement on write speed but I don't store large files to write. I use a conventional scratch disk 7200 rpm or a 300 gig velociraptor. I'm believe one should match up the SSD to all other devices it will eventually interact. So I don't have enough room to raid conventional hard drives nor do I want to purchase a third party pci card. For those of you suggesting otherwise, I'm spending my time towards myself and the normal casual user. Here's some points to consider:
1) I wanted to stop watch test only those real events I was familiar, finding that kind of data on the internet is basically nonexistent.
2) finding optimal raid block size is highly important, yet difficult to find. Probably because to test all the sizes would be a huge PITA. So my test is simple. Make the raid0, choose the block size, install the OS and test. Open DMG files off a second hard drive I used Seagate 7200 rpm and feel the time it takes to open the dmg and mount on the desktop. With a single SSD its quite fast, usually 1 second. In raid zero if the block size is chosen incorrectly then the time will become quite long, being obviously slower. I'm not the expert so I wouldn't recommend the correct size but I think 16k is good. 32K becomes slow.
3) Here are some stop watch times:
warm restart: single SSD 39.1 secs raid 0 47.6 sec (not worth the loss of one sata port)
install itunes 10.1.2 since this dmg does take a longish time: single SSD 1:47.0 min raid 0 1:27.6 (hardly worth the loss of one sata port)
use itunes to add to a very large music library including artwork: single SSD 1:44.1 min raid 0 1:44.1 min
using itunes to change equalizer to rock for the entire library: single SSD 1:17.9 min raid 0 1:16.8
How to make itunes numbers faster , instead of placing the music folder on a conventional hard drive, I used a 300 gig Velociraptor ; place the music file on the SSD then itunes will really fly faster BUT what's the sense for a one time shot. Why use up the SSD space for a couple extra seconds.
Lastly, I think raid 0 for conventional hard drives would be beneficial but once user gets into SSD the entire story changes. So I'm providing my experiences as a community service. I'm not looking for answer. I feel satisfied that I'm not going to raid my boot drive, a single SSD is fast enough to boot and launch applications. I'll store my data files on a scratch drive, a fast 7200 rpm hard drive wich matches up to my SSD drive. My SSD drives are X25M intel and they match well with fast hard drives. I don't need to raid hard drives because they'll write faster than the intel can do. 105 M/sec is good enough for me, the normal casual user.
My boot drive is 19 gigs used. I have the normal amount of Applications installed . My music, downloads and movies are on a 7200 rpm scratch drive. Attaching my firewire externals matches up with my internals for speed within reason. Someone write something positive so I can close this post out.

Answer 1

Another thought then I'm done with this thread.
I posted a topic about the Radeon 5770 or 5870 working in the 2008 Mac Pro. I read all the stuff on the internet that it is doable. Although it is there was not ANY comments about what if a person clean installs with that card installed prior to. I did that test and was not totally satisfied that its bullet proof.
I think that too many comments about things that supposedly work but will be unsupported needs to have more justification.
Except if you think about it, each time I post some simplistic real world test and make comments to how I'm satisfied with performance there are only a small less than handful of people who bother to return a reply. A conclusion then follows that what ever one digests out there on the internet needs to be looked at in depth with some thought. Not just acceptance that the data is all inclusive. I'd never post any data where I thought it was 100% correct when in fact there is variability.
A simple example would be do some research and try to find out exactly what programs use how many cores. There's some data but not enough to cover the large quantity of programs which exist for the Mac using either Snow Leopard or Leopard.
One of my pet peeves is to do research and go to some forum, out of this one and read all the responses to some persons query. Its ridiculous the amount of band width wasted with replies that don't come close to answering the initial question.