Introduction

Serial ATA is the latest hard drive storage technology to come along. It is a fairly significant change from the older ATA/IDE based standards even though it shares much of the same fundamental foundations. The technology offers us higher device bandwidth speeds, easy to configure drives and future upgradability. While the standard has been published for quite a while now, boards and drives supporting the SATA standards are only now becoming available in limited quantities.

To see what benefits the Serial ATA standard will bring to consumers, I decided to do a comparison of some of the first generation SATA drives available on the market and compare them to a high performance ATA/100 hard drive. Currently there are two consumer driver manufacturers, Maxtor and Seagate, that offer Serial ATA drives to the public and were kind enough to supply me with drives to review. Western Digital also produces the Raptor Serial ATA drive that I attempted to get as well, but this drive is marketed towards low-end server installations rather than desktop configurations.

Drive Specifications

Before we get rolling ahead into the benchmarking of the drives to see how they compare, lets take a look at the drives that are being tested for this comparison:

Specification wise, all three of the drives are very similar. The only real difference between the drives is in their storage capacities. This is caused from the different size platters used in the drives. The higher the platter capacity, the higher the inherent transfers speeds should be. Maxtor's drive is the newest drive using an 80 GB platter which should give it a significant boost and will be shipping shortly. Seagate's Barracuda V drive has been available for several months now and uses an older 60GB platter. The Western Digital Caviar is the oldest drive and uses a 40 GB platter size so overall its performance should be less than the newer drives, but its performance has always been strong.

The Western Digital drive uses the older ATA/100 standard that has a limited bandwidth of 100 MB/s compared to the Serial ATA interface with the 150 MB/s. Of course, even the fastest of today's drives can't burst traffic up to their bandwidth limitations for any sustainable amount of time.

The biggest thing to note is the warranties. Most manufacturers have reduced hard drive warranties to one year to cut costs. Western Digital has done this as well except for their Special Edition Caviar drives like the one in this comparison.

Physical Appearance

All three drives are standard 3.5-inch, 1-inch high desktop hard drives. The Maxtor drive has several exposed chips on the circuit board that the Western Digital drive hides on the other side of the exposed circuit board. The Seagate drive differs from the other two in that Seagate decided to fully encase the circuit board so there are no exposed circuits.

The primary difference between the three drives is in the connectors. The Western Digital drive has the standard 4-pin Molex power connector, IDE interface connector and the requisite jumper block for selecting master and slave configurations. The Seagate drive uses the standard Serial ATA connector block with the Serial ATA interface connector and the large 15-pin Serial ATA power connector. Maxtor also uses the standard Serial ATA connector block but adds an additional 4-pin Molex power connector to alleviate the need of a Serial ATA power adapter.

Test Configurations

Now that we have seen the drive specifications and what they look like, what about the test configurations used to benchmark the drives? Here is all the information with regards to the hardware that was used to run the benchmarks:

And the software relevant to the benchmarks:

The only hardware that changed during the course of the testing was the hard drives and the drive controllers. The SIIG SATA PCI is mentioned because I had intended to compare the results between the PCI based card and the SiI3112A on-board SATA controller integrated on the ASUS A7N8X Deluxe motherboard. It turns out the SIIG card uses the SiI3112A controller chip as well and the results were nearly identical for the benchmarks so I have instead chosen to only report the scores obtained from the on-board controller.

Next page > Benchmarks and Conclusions > Page 1, 2,

Previous Features