This is a cut from a review that was just so far removed from the actual point that it developed a life of its own. We are publishing it because you may find the information useful and it could save s buck.

Setup:

CPU: i5 3570

GPUs: IGP, AMD HD 7850 and nVidia 9800GT (just to see what MVP can do for an old card)

MEM: G.Skill 2400

Drives: 1x Mushkin Chronos 90GB SATA III SSD, 1x Seagate 250GB (3.5″) SATA III HDD, 1x Fujitsu SATA I 320 (2.5″) HDD, 3x Hitachi 500GB (2.5″) SATA II HDDs, 1x WD 500GB (2.5″) SATAII HDD, 1x ASUS E-Green Optical Drive

PSU: Thermaltake 750W 80+ Bronze

HSF: Arctic cooling ACFZ130

The first tests will be of the drives and the raid capability. HD Tune was used for basic setup testing because it is fast and does tell a story. It tells a story that dispels some things that are no longer true and that have been repeated for so long that some believe them to be Gospel. You can mix drive brands, sizes, and speeds in almost any combination on modern non proprietary controllers. This is not saying that it will work on an Adaptec Raid card (it has worked on silicon image cards) but I have been doing this since the advent of SATA and only had issues one time. As a matter of fact you can even mix a SATAIII SSD with a SATAI
HDD in raid one. You can take these results however you like but in my mind things are easier today than ever to just add a drive.

The following shots are left Fujitsu and WD in Raid 1 with Fujitsu and Hitachi in Raid 1 on the right. Do notice that the speeds are at SATA 1 and this is due to the delays imposed by the SATA 1 interface. These are still a fully functional Raids even though the brands, sizes and interfaces do not match.

Continuing with the mixing and matching we put different brands of the same size with the same interface together in Raid 1 (left) followed by the same drives in Raid 0 (right).

To take things a step further we mix all three on the left (three brands and two interfaces) in Raid 5 to see what we get compared to three identical drives on the right. Yes there is a performance hit but again this is due to the interface not the brand.

As if this was not enough testing we did more (yeah we love the flexibility of all those ports). We decided to see what some 4 drive mixed setups (all the same size and speed) would do. On the left is a Raid 10 setup followed by a Raid 5 on the right.

Finally we come to testing all drives together which should not work by the reckoning of some and be severely handicapped by the reckoning of others.

Above are five drives drives of three brands in two sizes with two different interfaces in raid 0.

So mixing and matching is doable on modern non proprietary controllers and though this is an Intel controller I have done much the same on AMD controllers. The manuals even say you can do it but yes it is recommended to use the same size and interface but that is recommended it does not mean it is a rule. Yes it is best to match sizes and speeds but it is not an absolute.