While surfing the net… elsewhere… we’ve seen some interesting comments from avid enthusiasts regarding the Maxtor Atlas 10K V vs. the WD Raptor WD740GD. In our recent review of the Atlas 10K V, graphs showcase the Atlas handily crushing the Raptor in both server as well as non-server (i.e., single-user) settings. When it comes to server use, these results are not unexpected- the Atlas benefits from a maturity both in its own design as well in the SCSI protocol’s command queuing implementation that makes it untouchable not just for the Raptor but also for the rest of today’s SCSI competition. When it comes to non-server scores, however, there is a catch.
Figures representing the Raptor in the Atlas article convey the SATA drive’s performance with TCQ enabled running on a Promise FastTrak TX4200. We chose this route as WD itself markets the Raptor as an enterprise-class rather than enthusiast-oriented offering. As demonstrated over the summer, while it helps server performance, SATA TCQ is detrimental to non-server use. Margins are much closer when the Atlas faces the Raptor with TCQ disabled. It’s scenarios like this where the SR Performance Database comes in handy. Here’s the relevant comparison:
Atlas 10K V vs. Raptor WD740GD w/o TCQ vs. Raptor WD740GD w/ TCQ
Bottom line- the Atlas generally tops the Raptor in non-server, single-user scenarios, but only by very slim margins. The Atlas 10K V currently remains the fastest 10,000 RPM drive one can buy for server use. For other applications, however, spending twice as much money to buy an increase in performance of about 5% is not the wisest use of hardware-enthusiast funds.
