Bilow,
Are you trying to hijack my post?
Here is an answer for your post. This was sent to me in a private communication.
It need not remain private...
Subj: Bilow's DI Post
Date: 98-06-08 22:25:28 EDT
I think this guy is missing something. The
supercomputer market is by no means dead. It is thriving as evidenced by the
success of IBM's Deep Blue and other massively parallel supercomputers.
The earlier supercomputer vendors were caught in
a recurring trap in the computer business, namely they were committed to their
architecture, and lacked the capital and perhaps courage to abondon what they
were selling to adopt a new software and hardware architecture. It is not hard
to understand how companies like DEC, DG, Wang, Amdahl, etc., etc. have gotten
hung up on the transitions. They have to abondon everything they have in order
to move forward. Their entire business base is tied to the
infrastructure they have created. The software architecture, their software and
their customer's software investment is tied to the architecture. Their internal
R& D talent and processes, the capital invested in plant and
inventory.ÿ To make the jump is virtually impossible, because it means
obsoleting everything they have built. Few companies have managed such a
transition successfully. The legacy overwhelms them, and they fall behind or
out. In the sixties IBM under TJ Watson Jr. bet the company, abondoning then
140x and 74xx series of computers and introducing the 360 series architecture.
It obsoleted itsÿ entire portfolio of software and turned its entire lease
base of hardware and software into instant scrap. The gamble paid off, and it
forced Univac GE and the rest of the seven sisters onto the scrap
pile.
What is interesting about this period in the PC
industry in not that were are facing obsolescense of the PC architecture, but
rather obsolescence of the business model that most of the major players have
built their business upon. Again the consequences are challenging. Compaq and
others must redesign their legacy business model, obsoleting their business
relationships and in effect cutting off their customer relationships as well,
leaving them open to being savaged by competitors who have created the new
business model. It is a bet your company decision. Not hard to fathom why Compaq
and others are temporizing. It is Dell that is likely to be the big winner here,
perhaps the only winner except in the very high end where the services
predominate.
One other long term trend in the industry is worth exploring.
The continued rapid growth of the computer industry. The costof a given level of
performance has dropped byÿ about 50% per year for the last thirty five
years. The software has not always evolved to take full advantage of the
performance available instantaneously, causing short term dislocations in the
demand for some vendors. But rarely has it lagged for more than a year. The
reason is that computers are tools for creating and communicating information.
As the performance increases so the potential for more, better faster
communications are enabled. There does not appear to be any natural limit to
amount of information or the sophistication of the processes to create store
analyze, display or communicate it. As good example is the WWW itself. UUNET
advertises that it is growing its hardware infrastructure a 1000% per year. As
we all know the demand of WWW services is consistently greater that all of the
various access providers and information providers can supply. This is the
nature of the information beast. The answer to the question when will we have
enough? The answer is NEVER. |
