Help me design a new hard drive testbed

[DougLite]

Well, after seeing some conflicting reviews of late, and various members dismissing reviews from other sites for a variety of reasons, I have decided "If you want it done right, you have to do it yourself." So, I am contemplating assembling a hard drive testbed.

I am on a really tight budget though, so that rules out a clash of the latest titans in the 300GB+ category. However, I do have a pretty large fleet of 160GB/7200/8MB drives in PATA (Seagate, Maxtor, WD) and could complete the field with entries from Samsung and Hitachi for less than $200. That would make a well rounded PATA test candidate battery.

On SATA, I should be able to pick up a WD740GD on the cheap as a payoff for recent 'consulting work.' The decision for what other SATA drives to include is a little harder, although I am leaning towards assembling another group of 160GB drives, although only Seagate has a native SATA drive at this capacity point - using an aftermarket PATA->SATA adapter would be the same method that the other manufacturers use to make their SATA drives at the 160GB capacity point.

The other hardware will most likely be (hopefully, it would be a great new gaming system) a DFI NF4 SLI-DR, 2GHz Winnie or Venice, 1GB of TwinMOS SP, and either a 6600GT or X800XL. I will also take my own advice and 'head over to the PSU forum' to see about a quality PSU.

As for software, it will be Windows XP Service Pack 2. However, what software should I allow to run in the background? None, in the interest of a pure testing environment, or a small selection of background apps such as anti-virus, instant messager, etc for more realistic tests?

I would like to test the following programs/games:

Steam
HL2
Doom3 (it uses a unique method for storing the game data)
UT2004 (LOTS of small files)
Either C&C Generals: Zero Hour or Warhammer 40,000: Dawn of War
SimCity4 (I have a city of ~1 million Sims that is quite disk intensive)

Also, probably some app tests:
Virus scan
Zip/unzip
Open to suggestions!

The overall goal here is to gather as much useful enthusiast info as possible, on an enthusiast platform, in an enthusiast software environment. Any suggestions/corrections towards this goal are welcome!

 

[unhappy_mage]

Do you have capability to do acoustic testing? I know it's not important to some people, but I prefer that my computer not sound like it's going to take off.

Unfortunately, I'm more interested in longevity of the drives than in performance, and there's really no way to test that without just running the disks for a year or two.

 

[DougLite]

I feel your pain about reliability measurement. There is simply no way to know. You do everything you can for clean power, have backups, etc, but it's never enough. Not even long term analysis of SMART data will give you any real clue, and you would need a huge sample to keep the impact of outliers to an acceptable minimum. I know better than to try and tackle reliability measurement.

Does anybody have a good way to measure sound, other than my ears being as fair to each brand as possible?

 

[Order]

Lets get a final answer regarding the benefits or handicaps of a 2-disk RAID 0 array for gaming.

 

[Bullitt]

IMO for background apps, just run any manufacturer supplied tools (mobo monitoring, soundcard mixer, vid-card tools, etc.)

Thats to keep it simple. Also, if I, as a super-duper power-l33t user who is trying to squeeze any performance out of my drives, I'm gonna turn off IM/Winamp/browsers anyways.

Basically, whatever you get "out of the box" and after all current windows updates are complete.

Acoustic tests would be cool, as well as perhaps surface temps of the drive, but to be honest, someone throws out a 33db number, another one throws out a 25db result, I'm confused. I cant associate those numbers with sound levels I have heard. Those are kinda meaningless numbers, but I generally pick lower ones if given same/similar (within 5%) performance specs. Objective and Subjective results would be nice (sounds like a spastic typewriter in my case, or sounds like a sleeping kitten) would be nice as well.

 

[unhappy_mage]

Does anybody have a good way to measure sound, other than my ears being as fair to each brand as possible?

If you were in MD, I could offer you access to a quiet room and decent recording equipment... but I don't think it'd work out too well.

Maybe you can post in the audio forum, and get advice on this? Maybe there's someone similarly equipped near you, or someone can recommend a cheap audio meter. Then build a soundproof box with power in it, put hdd in, put meter in. That would do it. Maybe something like this and a few of these?