A couple of thoughts on SUNW as gorilla...
Up front, I discount the Java theory ... not because it wasn't a very important Sun initiative, because it was, but because Java, per se, is a questionable profit center. Sun is almost profit averse in this area as evidenced by their having moved most of the Forte acquisition to iPlanet -- the EAI piece of which, in particular, is producing a very nice cash stream. They haven't moved the Java piece there yet, but then the real enterprise part isn't released yet and the feeling is that this will move to iPlanet as well when it comes of age.
I wish I had a nice simple theory of why, absent the Java, SUNW qualified as either a King or a Gorilla on clear terms. I don't. Frankly, for years, in an applied sense of what it is that I would put into my own customer sites, my bias was to HP since they seemed to understand business servers better while, for much of that time anyway, Sun was the choice for scientific sorts of applications. But, for what ever reason, Sun is clearly the dominant platform in serious Internet applications. This appears to remain true even while HP can claim such and such superior benchmark and IBM has some clearly interesting, as in argumentatively superior, servers based on their copper technology.
So, what's Sun's magic? Perhaps there is a real architectural superiority, regardless of whether the individual components are leading edge. Perhaps it relates to Solaris' effective utilization of the resources available. Perhaps it is nothing more than effective marketing. I'm not sure about that. What does seem clear is that Sun acts like a gorilla and gets away with it.
Might we have an interesting contrast in SUNW and QCOM? Sun acts like a gorilla and gets away with it, and yet it is hard to pin down why they should be able to do so. Qualcomm *is* a gorilla in very clear terms and yet seems to be stuck, for now, with deferred gratification. |
