Here's somewhat of a different tack on this issue. (And more than "enough
blue sky to patch a dutchman's pants.")
Personally, I'm of the opinion that somewhere in the 2005-2007 time frame,
if not before, we are likely to see the end of what we all call the
mid-range computer today. By then the market will have become much
polarized, maybe even completely -- into lots of "smart" things, which you
can't do much with beyond their "programmed by the manufacturer" abilities;
and really big boxes to run them.
What we all call a "Personal Computer" today, will be replaced by a real
"Home Computer" -- that is one that runs like a refrigerator and truly is a
general purpose home server. However, I would assume it to be several times
more powerful than something like an ES40 today, and it will be in the
$2-5000 price range. Retrofitting older houses will be expensive, but that
will just be rolled into the cost of new housing. Access to this box will
be via an assortment of Blue-Tooth and other wireless "smart things."
In the commercial market we'll see much the same thing happening.
Larry Ellison's Network Computers were ahead of their time but that will be
everybody's desktop and they will be running on 100meg (minimum) LANs
connected to massive ASP (application Service Provider) systems.
(This trend is often called "sever consolidation," and Oracle has actually
bet the farm on it.)
What I'm expecting is that there will be really rinky-dink machines on the
low end and really incredible Wildfire/Sierra class machines on the top...
The high end will be provided by Alpha and Power and the low end by AMD and
maybe Alpha and PowerPC -- but they're all going to look an awful lot
alike, and not much like the x86 world we know today.
There will be so much difference in computing coming in the next few years
that we'll look back on today just as today we look back on the PDPs that
gave birth to Unix... difficult to believe that they were THAT slow and
had only 4K of memory. (Yes, K, not meg.)
Intel, Sparc and Parc will be history. None of those architectures have a
long term future. And they barely have a short-term one.
Forbes (current issue) just did another piece on the fact that Sun is still
(a year later) having trouble keeping their high end Enterprise Servers
running. They're now claiming that it's cosmic rays effecting the cache
chips! It is undoubedly true, but it's an incredible, "oops."
"Windows" on the desktop is good for another 3-5 years, maybe. Gates knows
this, that why his push into the data-center and into the "set-top" world,
with the X-box.
Windows/Palm/or Linux for the "hand-held" stuff? Hmmm... one wonders if the
low end is in fact going to belong to Transmeta instead of AMD. We won't
know for a couple of years yet. [At 40watts it certainly won't be an IA64.]
But, despite all this, I'm stil pushing the Compaq folks to drive up Alpha
volumes by moving back to the desktop. Support for "real" graphics boards
and other consumer oriented things, like USB and Firewire, are essential,
along with applications.
But alas and alack, Alpha's problem is still Stealth Marketing. Many of us
hoped that Alpha and Tru64 would benefit from Compaq's marketing know-how,
but it seems that only the only people within Compaq who know what the
word marketing means, sell "Industry Standard Servers (x86)" or VMS.
Aha... but maybe that's Alpha's saviour -- VMS. VMS is now free to
educational sites. And VMS has a history of desktop support and clusters!
And they are the ones that taught SUN how to "market" into Education.
-- Give'em away so people grow-up on 'em, then they'll buy them when they
start their own businesses -- sad, but Palmer forgot that in trying to
"upscale" things.
IBM's new Power series will give Alpha a run for the High-end, but
fortunately, Alpha and Power share a lot of similar ideas as well as
Fabrication technologies.
Alpha and Tru64 volumes ARE going up, that means that prices should start
coming down. Hopefully soon.
Now, if I could only convince Compaq folks that lowering prices increases
demand.
There's a Unix knowledge contest starting next week -- check out the Tru64
home page (www.tru64unix.compaq.com), there are some nice prizes. Maybe,
this marks the beginning of some serious Alpha marketing.
Alpha has a future. IA64 blew it. And Bill Joy claims that he's terrified
of "nano-technology." I contend that's because he knows that Sun has no
place in that world.
Now if we could just find some applications...
In 2006 it will be an interesting exercise to see how much I underestimated
the changes.
Enough rampant speculation, it's obviously been a long day...comments?
--
www.tru64unix.compaq.com
www.tru64.org
comp.unix.tru64
T.T.F.N.
William H. Magill Senior Systems Administrator
Information Services and Computing (ISC) University of Pennsylvania
Internet: magill@isc.upenn.edu magill@acm.org
http://www.isc-net.upenn.edu/~magill/
_______________________________________________
Axp-list mailing list
Axp-list@redhat.com
https://listman.redhat.com/mailman/listinfo/axp-list
This archive was generated by hypermail version 2a22 on Fri Dec 1 08:00:06 2000 PST
Send any problems or questions about this archive to webmaster@alphalinux.org.