>Taking this into consideration, isn't the relationship between pc's
>(including servers) and embedded systems more like the relationship
>between automobiles and airplanes rather than barges and railroads?
MSFT is moving aggressively into the server market, and will succeed. But the production of servers is in the low millions versus annual PC production approaching 70 or 80 million, not to mention the installed base of 250 million PC for application software. Thus, I suspect the MSFT server business is not sufficient to provide adequate future growth, even with explosive growth of EIDs. And don't forget that a major attractiveness of Windows NT is its architectural compatibility with WinTel everything. EIDs need have nothing to do with WinTel, so without Windows clients, MSFT would have to face off against Sun on merit alone - and that would be a problem.
When the PC paradigm peaks, MSFT will be wounded, irrespective of what happens with servers -- unless it magically reinvents itself in time. History teaches that the reinvention is difficult, and usually happens only after the company in question capitulates and formally gives up trying to defend a fading paradigm. The problem isn't always that these companies don't see the shift coming. The problem is that at any point the potential new business (if they were to refocus) pales compared to their legacy business, and their internal culture is built around the legacy business. White papers circulated inside IBM forecast problems selling "Big Iron" as early as 1985, but the company resisted doing anything about it in a timely fashion. They didn't push PCs for fear of cannibalizing mini-computers and mainframes, justifiable since the PC business was peanuts at the time. They chose instead to let MSFT do it.
You might recall that I have seriously suggested that MSFT buy WIND - at any price - and operate it as a well-endowed, independent subsidiary. The fact that they haven't arrived at this idea independently, and acted on it, puts in question their ability to reinvent themselves in an orderly fashion - unless you believe multi-media content something-or-other will amount to an appropriate reinvention.
I'm not suggesting that MSFT should be shorted, even if it looks expensive. It may continue performing for years, and if it reinvents itself, maybe forever. What I am saying is that unless Gates begins to should show real vision, as the standard bearer of a soon-to-fade major computing paradigm, I think MSFT is risky.
Allen |
