said by drew :

I'm not trusting just what Matt is saying (though he's a pretty smart cookie) but more importantly, what AT is showing on their site.

It makes sense. I think there might be a higher rate of map loading (though the only game I really play is WoW) it's probably less of a perf. boost from putting it on a SSD. Now maybe if I put two SSDs into a RAID-0 something small would be there, but nahhhhhhhhhh

said by Matt :

I used to run RAID-0, all the way up to a 4-disk array. I've noticed significantly better performance from my single Velociraptor, especially with WoW map loading.

Keizer is right that certain things will be faster but I don't agree that game map loads will be as Anand disproved this in another RAID article and my own results bear this out. I noticed that if I copied large files around or was decompressing a large archive, RAID-0 was great. As you can see in the image above however, no significant difference and in this benchmark, RAID-0 was usually slower than a single drive. You could probably eek out better performance if you invest in a real PCI-E RAID controller with 15k SAS drives and on-controller cache though. But at that point the fastest SSD out will be cheaper and offer better performance.

Regardless, 100MB/sec for my Velociraptor plus ultra-low seek times results in a much better overall experience in Windows. Desktop usage is all about response time, not pure transfer rate. That's why SSD are great, as their access times are effectively 0ms and they have ridiculously high transfer rates.

said by drew :

Because it'd been a while since I'd looked at SSD prices and a second small sized HD was more affordable maybe?

said by howie :

I'll probably get a SSD at some point, but for now, I'm totally happy with my 7200.12 1TB SATA 3.0 HDD. It has ~130MB/sec. maximum transfer rates and averages about 103MB/sec. in a non-RAID configuration. My old 80GB ATA133 maxed out at about 40MB/sec. These new SATA drives are a LOT better than their old IDE counterparts.

said by drew :

And the machines are maintained identically, same software, same *everything*?

I trust benchmarks and people who do it for a living more so than anecdotal evidence.

said by drew :

I like their method better than yours when it comes to PC.

said by drew :

So in the area of PC benchmarking and tweaking, I shouldn't follow those who have tackled the idea/industry for years?

Uh. K.

said by drew :

Fantastic idea in the world of infinite resources (time and money, most notably.)

However when trying to figure out whether I should buy another Seagate SATA non-SSD drive to turn my system into a RAID-0 configuration or just purchase a SSD, I rely on others (plural!) to help me make the decision because my resources aren't infinite, but rather finite.

said by netgear3700 :

hard drive is Corsair P256.