[Eugene]

This Japanese conglomerate's latest drive builds upon the solid performance of the 7K400 with an all-new native SATA design that incorporates a second-generation 300 MB/sec interface and Native Command Queuing. The older 7K400 held its own for quite some time... how much better can the new Deskstar 7K500 do? StorageReview puts Hitachi's latest giant up against the competition!

Hitachi Deskstar 7K500 Review

[CougTek]

There is an error in this graph. It should be the 7K500 column that is shown with a different color, not the 7K400's.

[Atamido]

I immediately noticed the exact same thing. And here I was afraid to note it for fear of being a dork.

[Olaf van der Spek]

I keep wondering why it's so hard to get the NCQ implementation right.

[Eugene]

CougTek, on Nov 2 2005, 09:34 PM, said:

It should be the 7K500 column that is shown with a different color, not the 7K400's.

Fixed, thanks.

[Spod]

"The ultimate proof is in the pudding"? Shouldn't that read "The proof of the pudding is in the eating"?

I expect NCQ implementations will mature a bit, but it seems that the NCQ implementation can be classed under "firmware optimisations", and some manufactures will prove better at it than others, and some implementations will be better suited to particular uses.

It's actually nice to see a 7200 RPM drive catching up with the Raptor. Perhaps WD will consider this sufficient competition to drive them to update the Raptor design. And at least we no longer have to choose between capacity and performance (though price will, of course, be an issue).

[Eugene]

Spod, on Nov 3 2005, 09:32 AM, said:

I expect NCQ implementations will mature a bit, but it seems that the NCQ implementation can be classed under "firmware optimisations", and some manufactures will prove better at it than others, and some implementations will be better suited to particular uses.

This isn't an unreasonable assumption. Hitachi/IBM had ATA-4 TCQ quietly available over several generations of Deskstars- the prowess we see from the 7K400 may well be a result of all the practice they had over time. Hopefully they'll get better at the NCQ thing too...

[ehurtley]

Atamido, on Nov 2 2005, 09:55 PM, said:

I immediately noticed the exact same thing.  And here I was afraid to note it for fear of being a dork.

No need to worry, CougTek has the dork-ness covered for you. :-D (I kid, I KID!)

[Olaf van der Spek]

Spod, on Nov 3 2005, 03:32 PM, said:

I expect NCQ implementations will mature a bit, but it seems that the NCQ implementation can be classed under "firmware optimisations", and some manufactures will prove better at it than others, and some implementations will be better suited to particular uses.

Can it?
I can't think of a good reason for NCQ on to be slower than NCQ off.

[[ETA]MrSpadge]

Unimportant typo: on page 1 it should read "High-capacity..." for the MaXLine III.

Besides that, I wonder how important locality is for these amazing scores. Same for the WD RE2. With the prices the way they are now I'd rather be interested in 250 - 300 GB drives.

And last but not least: thx Eugene for having SR up 'n testing again!

MrS