Keeping my promise, I’m going to return to the problem of the support of Command Queuing technology by the WD740GD a.k.a. Raptor 2 hard disk drive. In our small investigation called WD740GD aka Raptor 2 Hard Disk Drive Review we put forward two suppositions:

1. The WD740GD hard disk drive may support Command Queuing (CQ), but the available Serial ATA controllers ignored Parallel ATA CQ, while the drive’s performance growth under high loads was due to the trivial sorting of the requests by the SATA controller’s driver.
2. The efficiency of PATA CQ may be so small that its effect is practically lost against the background of the speed gain the hard drive receives through the controller’s driver’s sorting algorithms.

Today I am going to continue investigating this problem from the point where I stopped last time. If you didn’t read the first article, here’s the gist of the problem: the WD740GD didn’t show any noticeable effect from the Ultra150/CQ technology (this is the name of the PATA queuing technology supported by Western Digital’s HDDs) with any of the available SATA controllers.

Promise SATAII 150 SX8 Controller

Not long ago we happened to get a sample of the Promise SATAII 150 SX8 controller into our greedy hands. This device supports Native Command Queuing, which is part of Serial ATA II extensions to the SATA 1.0 specification, and – that’s what we are interested in right now –Tagged Command Queuing (TCQ). This gave me an opportunity to try to “activate” the hidden capabilities of the WD740GD drive.

TCQ could be enabled or disabled for the WD740GD drive in the controller’s BIOS, but – a curious fact – I couldn’t turn TCQ on for the Seagate ST3200822AS. So, the controller differentiates strictly between NCQ and TCQ.

That’s how Promise’s controller looks like:

It’s based on a PSC81518 chip:

That’s an eight-channel PCI-X SATA controller with support of SerialATA II extensions to the SATA 1.0 specification. Let’s try the WD740GD drive on this controller now!