David -- thanks for your long and thoughtful post. You make many good points. I counter with the following arguments:
In each point you are correct. SGI faces a tremendous hurdle in trying to market an NT machine. However, it is also a tremendous market, and the high end is wide open. In particular the need for big servers that can manage video signals will explode in the next two years. SGI is sitting on top of that technology. I believe that conventional computer architectures won't be able to keep up and will turn into bottlenecks. Fancy SGI machines already pay for themselves, but SGI has simply been behind the 8-ball due to the fact that UNIX is being pushed out of the mainstream. (It will be interesting to see if SUN can reverse that process with DARWIN.)
Can SGI compete with marketing powerhouses like Compaq, IBM and SUN? Frankly, SGI has been a failure in the past in this regard. However, Belluzo is the perfect antidote for this deficiency. With the introduction of an NT machine and a $40 million AD budget, SGI is about to undergo a personality change. Hey -- could the advertising get much WORSE? Well . . . no.
Spin Off MIPS? Maybe so. Anyway, the cash value of that asset is one of the things that gives me confidence that SGI won't tank completely. And don't forget the $650 million in CASH -- a nice thing for an R & D company like SGI.
Is SGI the leading edge, or the bleeding edge? Well, SGI hasn't really been LOSING much money -- just growing less quickly. New markets, new challenges -- new expectations, maybe a new ball game.
Once last observation. I think the Intel world has basicly been a game of clone vs. clone. I expect SGI's new machine will be a few years ahead of the pack, and will command a suitably high margin. If so, then maybe SGI doesn't have to turn in to a "Gateway" or a "Dell." For the last couple of years SGI forgot Sutton's rule -- "Go where the money is." The money has moved into the arena of NT. Let SUN struggle against the tide, and we'll see where they are in five years.
--JD |
