Perhaps the biggest reason is that recent users of desktop PCs tend to demand more capacity to store DivX and MP3 files whereas web servers need better “multi-user” performances and reliablility. Some years ago, SCSI HDDs were ahead of IDE HDDs in terms of maxium space per unit. Now, SCSI has up to 300GB while IDE has up to 500GB. And more servers use IDE (especially SATA) HDDs for bigger storage requirement now. One of my “clubs” have 1TB web space and I doubt they use SCSI HDDs for most of the files. The second biggest reason is probably because data arial density must have something to do with rotating speeds. Here are examples:

Highest density at 7.2K RPM: 120GB to 133GB per platter (and will probably get over 150GB by next year)

At 10K RPM: 75GB per platter (Seagate 10K.7)

At 15K RPM: 36GB per platter (for Seagate 15K.4)

I think Raptors are more reliable than most other PATA/SATA HDDs. All reviews and user reports agree in that Raptors (especially the larger capacity model) running as RAID 0 array perform very well. Single Raptor or two Raptors as JBOD or independent drives are already a lot more expensive than all other SATA HDDs at 7.2K RPM so those who invest in Raptors are more likely to spend 2x more to use Raptors in RAID 0 arrays. I’ve seen many people since 2003 buying two Raptors at once on forums related to multi-processers and other topics. SCSI users (either desktop PCs or servers or workstations) are also more likely to spend x times more for RAID 5 or RAID 0 + RAID 1.

Pity Seagate and Hitachi are not making 10K RPM SATA HDDs even yet.


Thanks for answers :bow:

Raptors are about just as reliable as any generic HDD.

Raptors all the way!!!

noisier? well… yes a little… but… all WD are noisy compared to non noisy drives… this is of course a personal preference

speed boost… rdg said it best… ts all about the seek time baby…

when will you notice it??? not on day to day operations but encoding, rendering, and scanning will hit the spot (my norton does a scan in 7min vs the 22min it used to take)… then again day to day operations differ with the user… hence once again a personalized gain for some

Thats my 2 cents

that’s why the MTBF rating and warranty are longer than most any generic HDD…:rolleyes:

sounds like someone has raptor-envy :stuck_out_tongue:

5 years warranty isn’t very rare in case you haven’t noticed…

seagates and raptors are the only ones i know of…everything else is 3yr max and often just 1yr depending on the type/place of purchase (i.e. retail vs oem)…

JBOD presents the exact same “risks” as RAID-0, if one drive goes you lose the array. Of course, the true risks are no greater than with ANY HD. If the drive dies, you lose the data, period. If you’re appropriately backed up, nothing is lost in either case. So there’s no “reduction in reliability”, that’s just not the case.

All drives die.

The Raptors are rated for over 1 million hours. All drives carry the same risks of a platter failure, but apart from that you can expect many years from a Raptor.

The topic here is not RAID, but HD’s in general.

As to the differences in access times (not seek times), you most deffinitely do see it on desktops with the Raptors. My boot time is 1/2 what it is on any other drive with the Raptors. Virus scanning is about 1/3 the time, and multitasking is equally improved. The Raptors will burn 2 different DVD’s at 8x PCAV simultaneously, try that with any other drive, and then do some surfing and a couple other things too.

That’s why I’d like to see some real world testing of the 15000 RPM drives to see if their access times are any better than the Raptors. All in all, HD’s spend more time seeking and accessing than they ever do transferring.

No rdgrimes, JBOD --> Just a bunch of discs
Which means that they doesn’t depend/rely on each other.

umm, no… :disagree: :rolleyes:

JBOD can be thought of as the opposite of partitioning: while partitioning chops single drives up into smaller logical volumes, JBOD combines drives into larger logical volumes. It provides no fault tolerance, nor does it provide any improvements in performance compared to the independent use of its constituent drives. (In fact, it arguably hurts performance, by making it more difficult to use the underlying drives concurrently, or to optimize different drives for different uses.)

in proper JBOD spanning, the HDDs are reliant on each other still…

Absolutely. It’s amazing how much people who don’t use RAID seem to know about it, isn’t it? :rolleyes:

JBOD is just another array that spans multiple discs. If one disc dies, the whole array is lost.

I’ve tried to make this point before, because this discussion keeps circling. In the end, several responses here went off topic.

We are now discussing which HDD setup is better, the original question is comparing and discussing a Maxtor vs a WD Raptor… and which would you prefer, perhaps a reason why. Lets try to revive ourselves on track here again…

(Just a Bunch Of Disks) A group of hard disks in a computer that are not set up as any type of RAID configuration. They are just a bunch of disks."
What you’re looking for is spanning which also is called JBOD.

from the link previously referenced/quoted (i trust over when discussing HDDs):

If you have some disks in a system that you decide not to configure into a RAID array, what do you do with them? Traditionally, they are left to act as independent drive volumes within the system, and that’s how many people in fact use two, three or more drives in a PC. In some applications, however, it is desirable to be able to use all these disks as if they were one single volume. The proper term for this is spanning; the pseudo-cutesy term for it, clearly chosen to contrast against “redundant array of inexpensive disks”, is Just A Bunch Of Disks or JBOD. How frightfully clever.

and per xtacydima’s request to keep this on-topic: i would choose raptors over 15k scsi simply based on the price-per-GB breakdown…

Kenshin, are Raptors a loss-leader for WD? i would assume given their capacity for making drives, WD’s margin on raptors is more than adequate to make up for any increased production costs (which i’m guessing are minimal)…

Well, I can quote numerous of source about this confusing naming… can’t we just drop it?

The Maxtor claims seek times “as low as 3.0ms”, but does not provide a full set of specs (at least not readily available on the web site). Since seek time is a relatively meaningless spec, it’s hard to draw any conclusions about how the Maxtor would compare to the Raptor. Which is why I’d like to see real words tests on it for access times and multitasking tests.

The Raptor specs offered are much more detailed, but still only relative seek times.

The Maxtor is rated up to 98MB/sec sustained transfer, which is not too exciting. Better than the Raptor (`75) but probably not too noticable in most cases.

I don’t think I’d turn down a free Maxtor though, but I doubt I’d feel much difference over the Raptors.

For the uninitiated: “seek time” is the time it takes for the head to move to the location of the requested read or write. Add this to the time it takes for the platter to spin around to the requested sector and you have access time. Higher RPM should lower the access time, but in the real world it does not always do this. The biggest impact of the higher RPM is in the sustained transfer. But unless the data being transferred is continuous on the platter, you may not even see that benefit.
So the drive with the fastest access times always wins in real world tests, because we all have fragmented files and multiple tasks going on. When multitasking or dealing with a fragmented file, the drive with a faster access time can even beat a drive with faster transfer rates.

You’ll also notice in the specs that drives function very differently, speed-wise, when writing than when reading. Speeds can be very different. Personally, I’d prefer faster write speed and slower read speed over the opposite, but you rarely see this. When transferring from one drive to another, it’s always the writing drive that is controlling the speed of the transfer. When reading and writing on the same drive, it’s the access time that will make it seem slow or fast. (and the size of the cache).

If I could choose, I would have chosen the Maxtor 15k SCSI drive. Because SCSI is reliable and fast. But the Raptors too ! The 15.000rpm drives have even higher transfer rates and access times. Then again, it depends on what you use it for. If it’s for desktop, go for nice 16MB 7200rpm drives or raptors. For workstation, go for raptors or 10k or 15k SCSI drives. Servers: SCSI only.

I use my PC as desktop/workstation (Photoshop) and I like my PC nice and smooth. So I went for 2 raptor 36GB in RAID0 where Windows, my games and music are installed. I have two 160GB Maxtor DiamondMax Plus 9 drives, mainly used for storage. But on one I have the swapfile. Why ? Well, when loading a game or working with Photoshop, stuff is writen to the swapfile. Yes even if you have 1-2GB of RAM. Putting the swapfile on another drive solves the problem of the big delay caused by reading stuff from the game and writing it to the swapfile on the same drive. What you get: 1 cycle is used to read a block of data on the drive, another cycle is used to write it on that drive. Also the overhead caused by the movement of the head that needs to be done. When you put the swapfile on another drive, it can read and write at the same time, without repositioning the heads everytime.

And because I’m running RAID0 and I have lost enough data by crashing IBM drives, I have a backupdrive of 250GB which is just principally a copy of the important stuff of the other 4 drives. SecondCopy is the software which does this every 2 hours automatically in the background. No need for expensive RAID5 controllers and at least 3 drives. And after 2 years about 12 hours a day running time, it hasn’t ever let me down while keeping the performance satisfying :slight_smile:

Maybe another thing to consider in choice is heat production: a 15k rpm drive produces more heat than one at 10k rpm. This can require another fan and then more noise.