What RAID does depends on what version of RAID you use. Quickly; RAID 0: Data striping. Two drives in parallel, with the data spanned across both. Pluses: almost double the read/write bandwidth, so much faster read/write actions. Minuses: Exactly -no- fault tolerance. You lose a drive, all the data it gone, as it spans both drives and is synched. RAID 1: Drive mirroring. Just that. Two matching drives that are written to simutaneously. Pluses: Data safety. One drive blows, the other takes over automatically. Minuses: No improvement in access speeds. RAID 0+1: Combines the two. Striped drives for speed increase, with mirroring so that if one array goes, you have another. Pluses: Best of both basic types. Minuses: Expensive; you need 4 drives minimum for -each- array. Plus you have power costs and head issues to contend with. You also need to keep a spare set of drives handy, as you need to replace the failed array and recreate is as soon as possible. Too many figure they have time, and if you build the array with all new drives, then they are all equally old. When one goes, the other 3 won't be far behind. Many other types of RAID. Google it. SSD's are the new thing. There is some speed improvement, but you have to see just what size hard drive they are testing it against. A sata 40 gig drive will smoke an SSD, where and SSD will stomp all over a terabyte HDD in benchmark tests. But the one thing you do =NOT= want to do is trust perishable data to an SSD. It is nothing more than flash memory, the same thing that is in every thumb drive out there. And there are physical limitations to the number of times you can write to a NAND gate before it fails. It's physics....writing to a NAND gate and changing it's state causes atomic level damage to the actual substrate structure. Currently, you have around 1 million writes to a NAND before failure to hold state occurs. And with SSD's structuring, if a bit gate fails, the associative byte is gone. SSD's just aren't old enough yet to know how the failure pattern is going to be; steady erosion of space, or large clumps suddenly dying (and a million writes sounds impressive, but consider how many times a second Windows hits the swap file....). As long as you have the OS disc, putting the OS on an SSD could be a good thing....but I would put -only- the OS, and any codecs and code snips that =had= to be with the OS with it. All other apps and certainly your perishable data should be kept on proven hard drives....at least until SSD technology has been real world proven regarding durablility and lifespan...