When i switched (temporarily) from my 15k scsi drive to a still very fast 10k scsi drive, the difference was very noticeable. Opening a program or file on the faster drive was MUCH quicker, but obviously gaming performance was either not affected or barely at all. The difference might also be due to in part the difference in SCSI standards (Ultra320 and Ultra160)
And the noise is often exaggerated - my 15k drive is only 1 dBa louder than my 10k
SCSI can be really hit or miss and if you’re not using something fairly close to the most recent line of SCSI drives you’ll quickly find that performance of newer 7200RPM drives will have eclipsed the older SCSI technology.
Like for example, I got a Fujitsu MAU3036NP which is one of the fairly current 15k scsi models. It has a single 36gb platter and spins at 15k rpm, has a transfer rate of 100-60MB/sec. I went on ebay and also bought a few older 15k fujitsu SCSI drives to round things out. I got two additional 36g models, but these ones were MAM series drives which were a 4x9.1g platter design.
The two didn’t even compare. The MAM drive pretty much picked up where the MAU drive left off, having a transfer speed of 60-40MB/sec or so and all in all was really not that fast, seeming even slower than the 36g first gen raptor I also had in my box at the time. Beyond that, the MAM drive was so loud it was retarded. It wasn’t loud the same way a loud fan is loud, it had a unique high-pitched whine that would seem to break through any and all barriers; I could hear it clear across my house.
Another example, I have a dual xeon rackmount server that I’ve used for hosting games at LANparties in the past. I got the server 2nd hand, and it had two 10k 9.1gb drives and a 10k 18.2gb drive. I was just running a Counter-Strike: Source server on it. The server files were on the 18.2gb drive at first. I noticed it was taking a long time to switch maps. I eventually changed it over to using the two 9.1gb drives in raid-0 to try to speed things up. The maps would load so slowly that people were actually complaining. We then switched over to using a laptop with a 80gb 5400rpm drive in it and the problems more or less went away. I ran hdtach and hdtune on the server and was totally taken back by just how slow the SCSI drives in there were.
So yeah, my point is that it’s almost worthless to go with older SCSI tech, and it will really not be representative of what a current 10k or 15k drive would be. I’m not surprised in the least bit that you found your smaller, older 10k drive to be slower than your newer, larger 15k drive. That certainly does not mean, nor would it really make any sense to say that a drive with a faster RPM will automatically be, or even “feel” faster. Building on that, it’s more difficult to make a definitive conclusion as to which drive will be faster when comparing a current 7200rpm drive with what is now a fairly old tech 74g raptor; certainly much more so than if you were comparing against a current 16mb cache raptor.
