[NEWS] - StorageReview reviews Maxtor's MaxLine III.

[CougTek]

In this third-gen unit, Maxtor integrates three platters to yield a flagship 300 GB capacity as well as a slightly smaller 250 GB unit. The firm specs seek times at 9.3 milliseconds.

The MaXLine III brings significant changes to the table. Previous SATA MaXLine products (and DiamondMax products, for that matter), were designed with legacy parallel operation and retrofitted for the SATA interface through a bridge chip. The MaXLine III represents the firm's first from-the-ground-up SATA product. As a result, the drive does not feature a standard 4-pin molex-style power receptacle like earlier Maxtor SATA offerings did, instead relying exclusively on the 15-pin SATA standard. Note that a parallel ATA version featuring standard 40-pin PATA and 4-pin molex connectors will also be available.

Along with this native implementation, Maxtor has incorporated some second-generation SATA features, most notably native command queuing (NCQ). Command queuing allows a drive to intelligently reorder requests to minimize seeks distances and rotational latency. Though CQ has been implemented in the SCSI world for years, it is just now reaching ATA shores. When paired with an appropriate controller, NCQ-enabled drives potentially enjoy significant increases in performance under heavy-depth, highly-random operation.

The MaXLine III also features a 16 MB buffer, the first increase since Western Digital upped the bar to 8 megabytes nearly three years ago. A large buffer and its accompanying firmware is much the opposite of command queuing- read-ahead and write-back optimizations shine in highly-localized scenarios.

What an astonishment! SR posting the evaluation of a drive the same day as it is announced. I almost kicked my jaw when it felt on the floor.

News source

 

[Santilli]

3 YEAR WARRANTY...

S

 

[cquinn]

Figures... I just bought a WD2500JB the other day.

 

[ddrueding]

oooh......16MB buffer...drool

Can't wait to read the review.

 

[LunarMist]

It is interesting. Access times suck big time. Is that indicative of future, drives with denser platters, or will perpendicular drives fare better? It is also interesting that it does well in SR test suites. I think they are affected by cache more than SR would like to admit. Maybe it is time for some newer benchmarks.

 

[blakerwry]

maybe real life is affected by cache as much as the tests... hard to tell.

 

[Santilli]

I suspect cache, algorithims, seek times, and most important, PCI processor speed and interface speed, are very important.

s

 

[BSD]

Does anyone know if the NCQ will work in linux???

 

[blakerwry]

Don't we have to wait for S-ATA to work first? hehe..

really, S-ATA support in linux is basically beta status. Since linux has TCQ support it shouldn't be a problem to implement NCQ support once basic S-ATA support is there.

 

[BSD]

didnt know there was poor support for sata...might go with ultra 320 then [=_-]...