In Practice: No Apparent Problems
Experience has shown that, even when new technologies sound promising, it can take weeks or months before the first, minor bugs come to light. All of the serial ATA devices we used for this article ran like clockwork, although we did notice a few strange occurrences that can only be explained by the fact that we were testing pre-production models.
There were a few raised eyebrows the first time we tried to connect the Western Digital Serial ATA hard disk to our HighPoint RocketRAID 1520. We were unable to connect the power cable and the data cable at the same time. The problem might have been the data cable that came with the controller, because its connectors were 2 mm wider than the cable supplied by Western Digital, or, more likely, that the inputs on the hard disk are somewhat too close together. To be fair, we were dealing with prototypes, and we don't expect to see this kind of problem on production models.
The second odd experience occurred when we were carrying out benchmark testing and gently touched the hard disk on the table. It seems that this brief contact was enough to cut power to the disk, because the drive cut out and restarted. We couldn't believe it, but managed to repeat the trick a few minutes later.
The dips in transfer speed are clear enough. They coincide with the system reboot caused by our shaky contact.
We were also able to reconstruct this with the HighPoint controller. The plugs on the cable supplied with the prototpe drive and the two cables supplied with the controller were a very loose fit. We imagine that there would be tremendous problems if production versions are the same as these. Our sample Serial ATA connectors were loose enough to fall out simply from moving the PC. Here's hoping this is sorted out by the time the production versions arrive.
