Author: Mike Karp, Senior Analyst, Enterprise Management Associates
There cannot be much doubt that Serial Attached SCSI (SAS) is ready for prime time. The technology is new but evolutionary, and therefore it doesn't require much of a stretch for system builders to understand it. It is a proven technology: we have seen SAS work in the lab, at plugfests and in the field. It's hot pluggable, has a roadmap that will take it to 4X its current speed, it's cheaper to build than parallel SCSI, it plays well with the cheaper SATA technology, and it's highly scalable.
Additionally and perhaps most importantly it arrives on the scene (not coincidentally, of course) at a time when parallel SCSI, which has played a major role in protecting IT data since the 1980's, has reached what may reasonably be described as its end of life. Because something is a technical success however, even if the timing is right, does not necessarily guarantee that it will win in the marketplace. Remember the Betamax, which by most accounts was technically superior to the VHS videotape technology that eventually defeated it? Remember mini-super operating systems like VMS and PRIMOS? Many readers of this newsletter could make a good case for the fact that both were far superior to Unix in many ways, and yet both fell by the wayside.