Is the Raptor ever going to get NCQ? (not TCQ)

[Karma]

They've been out for a while, you'd think the whole line would've been updated by now.

[mjz_5]

i was just thinking that.. I hope so, especially after reading the article at anandtech.

[serbiaNem]

That's a good question, another question is whether motherboards with get TCQ support.

On paper it seems TCQ is faster than NCQ. http://www6.tomshardware.com/storage/20041116/command_queuing-01.html

I know Tom's has biased CPU reviews, but for other hardware they are fine.

[DougLite]

WD would have to completely redesign the controller board on the Raptor and develop our outsource a native SATA logic set before they could implement NCQ (or any of the other SATA-II features for that matter) as it is actually a PATA drive with a bridge chip, like all other WD SATA drives. I'm sure it will happen eventually though. Ironically, WD claims that SATA-II will overtake SAS - when they don't have a SATA-II (or SAS) drive on the market!

[UICompE02]

Ironically, WD claims that SATA-II will overtake SAS
That would be interesting, since SATA II doesn't support many fun and useful SAS features such as wide ports, expanders, dual porting, 16,000+ drives in a domain, etc. I wonder why WD doesn't just make SAS drives if they're so worried about it!

[FunkStar]

Does running a raptor/drive without NCQ through a PCI-X raid-controller with NCQ give the same effect as the drives themselves having it? cause that would be one way around the wait if it works that way (i doubt it, but never know)

[dolphumous]

74GB Raptor w/o Tagged Command Queueing

vs.

74GB Raptor w/ Tagged Command Queueing

http://www.storagereview.com/php/benchmark/compare_rtg_2001.php?typeID=10&testbedID=3&osID=4&raidconfigID=1&numDrives=1&devID_0=259&devID_1=267&devCnt=2
----------
What's TCQ and shit?

http://www.storagereview.com/articles/200406/20040625TCQ_1.html

TCQ vs NCQ

http://www.tomshardware.com/storage/20041116/command_queuing-01.html
----------

EDIT: I'm not sure why I just put all that there...lmfao...

[UICompE02]

Does running a raptor/drive without NCQ through a PCI-X raid-controller with NCQ give the same effect as the drives themselves having it? cause that would be one way around the wait if it works that way (i doubt it, but never know)

I wonder if that's really possible. NCQ is effective becuase the drive hardware/firmware knows the physical layout of where the sectors are on the drive. The controller can't tell as effectively where each sector is since requests for reads and writes are using the LBA (logical block address), which is not the same as where the data is stored physically on the drive.