Purchase of SGI by Compaq?

[James Helman]

The San Jose Mercury News article was taken from the New York Times.
That the SJMN needs to rely on a newspaper in New York for covering
Silicon Valley business never ceases to amaze me. ;-{

A couple sentences in last paragraphs say a lot:

	"Silicon Graphics last week adopted a so-called poison pill in
	the form of a preferred-share purchase right to be distributed
	to its stockholders in the event of any proposed takeover."

	"But there have been few successful mergers between technology
	companies, and the free-wheeling corporate culture of Silicon
	Graphics might clash with that of Compaq, which is more
	staid." [NYT version only].

With good reason, I doubt SGI is eager to be acquired.  But I think it
could be a useful alliance given another form of cooperation.

Why?

1) MIPS-based Unix workstations are getting trounced in terms of unit
volume by SPARC-based ones.  This affects what ISVs port to, which in
turn affects platform sales, starting a snowball effect.  A platform's
software base matters more as graphics moves into mainstream
applications.

2) The graphics playing field will be getting closer in performance in
the next 10 years.  We don't yet have enough performance yet for most
current applications, let alone things like VR, but by then, we may.
If so, given the choice between a PC successor, a Sun, and an SGI, all
with graphics adequate for 90% of applications, but with very
different general software bases, who wins?  Volume or innovation?
History makes one cynical; I'd put my money on volume.

[At Vis '90 last October, I asked Jim Clark (SGI founder) and Bill
Poduska (Stellar founder and Stardent hitech mergerer) a question
along these lines.  Clark said something to the effect that SGI was
going for innovation, which would lead to the critical size that a
workstation company needs to survive.  After that, Poduska didn't even
try to answer.  This is one place where size matters.]

3) Things are really shaking now.  DOS can't hold on forever.  NYT
rumors of a Microsoft, Compaq planning a MIPS R4000 based OS/2
machine.  SCO on a MIPS.  SGI's move into the PC world with a board
set.  Sun's lagging behind SGI in graphics performance.  It's not
clear what will happen; but transitions make for opportunities, and
Compaq and SGI could complement each other well.

Interesting times.

Jim Helman
Department of Applied Physics			Durand 012
Stanford University				FAX: (415) 725-3377
(jim at KAOS.stanford.edu) 			Work: (415) 723-9127