Western Digital VelociRaptor 300GB

"Clever girl..."

- Robert Muldoon.

Perhaps all that Western Digital need do is release a special edition and enjoy the increased sales by Jurassic Park geeks worldwide.
Of course, that would mean that we would end up seeing 2 more versions with everyone attempting to forget about the last... :)

"SHOOT HER!"

- Robert Muldoon

Wondered when you were going to get round to this awesome drive, been using one for some time now, almost since release. It's essentially my Steam directory ¬_¬

I feel I let myself down here - this review needed more JP quotes

I was musing about something to do with Unix systems.

JP \o/

http://images.bit-tech.net/content_images/2009/03/western-digital-velociraptor-300gb-review/velociraptor.jpg

I have 2 of these and they are awesome.

damn i want an ssd, i keep telling myself to wait a year or two but i want everything i want

On the first page you say that the 334GB platters of the 1TB drives have a much higher platter density, but I beg to differ. Assuming that the platters are 3.5" and 2.5"* respectively with a 1" hole in the middle - though I realise it is a bit smaller for 2.5" drives - we have the area of the 3.5" drive is 11.25Pi sq inches and the area of the 2.5" drive is 5.25Pi sq inches.

This gives a density of 29GB/Pi GB/sq inch for the 2.5" drive and 30GB/Pi GB/sq inch for the 3.5 inch disk. This is a very small difference, especially when you consider that the spindle hole in the middle is smaller on the 2.5" disk, which will probably mean the platter densities are the same, if not slightly higher on the smaller disk.

Another issue is the 16Mb of cache the VelociRaptor has. Well great, my two-and-a-half year old drives have 16MB caches. The 1TB drives often have 32MB caches.



*if I have used the feet shortcut instead of inches, I apologise

edit: it has been pointed out that the smaller spindle on the 2.5" drive will actually reduce data density :o

Originally Posted by Cupboard
On the first page you say that the 334GB platters of the 1TB drives have a much higher platter density, but I beg to differ. Assuming that the platters are 3.5" and 2.5"* respectively with a 1" hole in the middle - though I realise it is a bit smaller for 2.5" drives - we have the area of the 3.5" drive is 11.25Pi sq inches and the area of the 2.5" drive is 5.25Pi sq inches.

This gives a density of 29GB/Pi GB/sq inch for the 2.5" drive and 30GB/Pi GB/sq inch for the 3.5 inch disk. This is a very small difference, especially when you consider that the spindle hole in the middle is smaller on the 2.5" disk, which will probably mean the platter densities are the same, if not slightly higher on the smaller disk.

Another issue is the 16Mb of cache the VelociRaptor has. Well great, my two-and-a-half year old drives have 16MB caches. The 1TB drives often have 32MB caches.

*if I have used the feet shortcut instead of inches, I apologise

Excellent point, well made.

By our own calculations, the theoretical data density (considering a 1" motor on both drives) is, as you say, remarkably similar - 36.381 GB/Sq Inch for the Raptor vs 37.9 GB/Sq inch. However, Western Digital don't directly publish the figure on data density, so we'll have to go on these numbers. However - the Raptor probably doesn't use the more efficient perpendicular recording method that's allowed larger capacity drives to keep up performance wise, which is likley why the performance wasn't that far ahead of the competition even despite the 10,000RPM jump in drive speed.

You seem so impressed with the speed of these SATA drives which would seem that there isnt anything else faster.

Why do you not review SCSI drives? I mean SCSI has been around for years and years and have been so much faster thatn IDE, SATA, SSD and still are.

So when will you do SAS drive test and compare all?

Greetings!

I have one of these, the 150GB little brother, it is great and I am glad to have purchased when it came out. I use it for OS and applications.
That said, I think I wouldnt buy one today, there are some decent SSDs out now with an acceptable price.

One thing to note about the VelociRaptor is it's markedly quieter than previous raptors and indeed most desktop hard drives.

As for comparisons to SCSI drives. It's quite simple. They're damned expensive and a pig to setup. Just not a viable option for most consumers. Especially as just one drive gives enough capacity for your main system. SCSI only makes sense if you want speed and high capacities.

One final point. Seemingly like many things, these drives have actually gone up in price recently. These were £180 on the day of launch and dropped to £150 soon after and now they're £200+. Dunno quite what to make of that fact but there it is.

"At £200 for a 300GB drive (£1.50/GB)"

umm thats not quite correct, its 1.5GB/£

Originally Posted by mikeuk2004
You seem so impressed with the speed of these SATA drives which would seem that there isnt anything else faster.

Why do you not review SCSI drives? I mean SCSI has been around for years and years and have been so much faster thatn IDE, SATA, SSD and still are.

So when will you do SAS drive test and compare all?

I think that would be great too. But most people don't even know what SCSI drives are thou.
It's sad...... :(

Originally Posted by djDEATH
"At £200 for a 300GB drive (£1.50/GB)"

umm thats not quite correct, its 1.5GB/£

Fixed

Sorry guys - no plans to test SCSI drives in the future - they're pretty much exclusive to the enterprise environment and even then they're not too widely used. We like to cover some niche hardware, but in this case we're not sure there's much point.

Originally Posted by Jamie
JP \o/

http://images.bit-tech.net/content_images/2009/03/western-digital-velociraptor-300gb-review/velociraptor.jpg

http://imgs.xkcd.com/comics/velociraptors.jpg

Originally Posted by mikeuk2004
You seem so impressed with the speed of these SATA drives which would seem that there isnt anything else faster.

Why do you not review SCSI drives? I mean SCSI has been around for years and years and have been so much faster thatn IDE, SATA, SSD and still are.

So when will you do SAS drive test and compare all?

The cost of SAS drives would be horrible considering nearly all if not all mainstream motherboards don't have SAS connectors on them so you would have to buy a decent SAS card which could cost you between $80 USD up to over $1000 USD depending on how may drives you want to connect to it and the different RAID options you want. Not to mention even price of the drives... $180 USD for 147GB drive and up to $340 USD for a 300GB drive.

So, even if you wanted something to just compete with a single raptor it would cost you around $430 USD PLUS TAX! Not to mention you would need to buy the SAS cables for them but that's chump change.

Oh, one other thing SAS drive are loud as hell. They make 80mm Delta case fan sound quiet.

Good points Cupboard but:

Originally Posted by Cupboard

This is a very small difference, especially when you consider that the spindle hole in the middle is smaller on the 2.5" disk, which will probably mean the platter densities are the same, if not slightly higher on the smaller disk.

You were mostly correct in your argument til here, if the spindle hole (or motor) was smaller on the 2.5" disk, that would increase the area of the actual platter, therefore decreasing density. So the reality would be the 2.5" drive would not compare as well density wise if you took into account the smaller motor.

Originally Posted by mikeuk2004
You seem so impressed with the speed of these SATA drives which would seem that there isnt anything else faster.

Why do you not review SCSI drives? I mean SCSI has been around for years and years and have been so much faster thatn IDE, SATA, SSD and still are.

So when will you do SAS drive test and compare all?

The thing about SAS drives - they suck for desktop use.

The firmware is tweaked for high IO and random R/W, neither of which feature heavily in a desktop environment (yes, even if you multitask lots).

Consumer drives are the best option for consumers, funnily enough.

wow Im so glad I went for two 1.5tb Seagates because by those results much better buy ^__^

It's a shame you didn't try 2 x Samsung F1 in RAID0 as nowadays most motherboards have the ability to do that and as the drives are over twice as cheap a comparrison would have told a lot......

Originally Posted by mikeuk2004
You seem so impressed with the speed of these SATA drives which would seem that there isnt anything else faster.

Why do you not review SCSI drives? I mean SCSI has been around for years and years and have been so much faster thatn IDE, SATA, SSD and still are.

So when will you do SAS drive test and compare all?

Agreed. I used to (A long time ago now, when XP was new) boot my system of a 36GB 15k RPM SCSI drive on a U160 bus. That system booted extremely quickly, and while it didn't have the blistering fast data transfer of SATA, it's main rival at the time was DMA133, which it beat with ease.

I was looking at prices on SAS disks here (In Canada) - a Fujitsu 15k 147GB 16MB Cache SAS drive comes to $179. That's not a bad price, and certainly beats out prices for SSDs, but how does the performance stack up? And yes, I know they're noisier than 7200RPM or 10k drives, but it's a higher pitch whine, almost in the inaudible range.

Oh, and "Random R/W"? Sounds like a system boot if ever I heard one. Being optimised for Random R/W means low seek times.

Or, if you can find 'em still, how about an older, cheap U320 15k SCSI drive, on a cheap Adaptec U320 card.

Originally Posted by Anakha

Oh, and "Random R/W"? Sounds like a system boot if ever I heard one. Being optimised for Random R/W means low seek times.

It sounds like it, but its not. Random R/W means right across the whole disk, boot time activity is pretty localised.

Seek times make surprisingly little difference in desktop systems.

Originally Posted by Vimesey
Good points Cupboard but:



You were mostly correct in your argument til here, if the spindle hole (or motor) was smaller on the 2.5" disk, that would increase the area of the actual platter, therefore decreasing density. So the reality would be the 2.5" drive would not compare as well density wise if you took into account the smaller motor.

oops!

Any of you guys ever reviewed one of Hitachi's 15000RPM drives?

Originally Posted by Baz
Fixed

Sorry guys - no plans to test SCSI drives in the future - they're pretty much exclusive to the enterprise environment and even then they're not too widely used. We like to cover some niche hardware, but in this case we're not sure there's much point.

shame, im just curious how SAS compare to everything else.