Subject: Re: 1980s the "golden age" of super-computing
From: mash@mash.engr.sgi.com (John R. Mashey)
Date: Nov 20 1996
Newsgroups: comp.sys.super,comp.arch

In article <32831A20.378A012D@clarkcom.com>, Glen Clark
<glen@clarkcom.com> writes:

|> Curious how different viewpoints can emerge from looking at
|> the same picture. The recent article on SGI in Business Week
|> seemed to imply that the Cray product line would keep food on
|> the table while the SGI product line tried to recover from
|> being perceived as a motion picture-dominated culture.

Having been away for a month, I missed this, but in general,
without criticizing Business Week directly, I'd suggest that relying
on such [fairly general press] magazines for deep insight on high-tech
companies is usually doomed to failure.

1) There are some very good and knowledgable people in the press,
who keep at some area long enough to be quite insightful.

2) A few magazines are consistently good, and actually admit mistakes
when they make them. Some aren't.

3) There are some reasonable people in the press, who try hard, but:
	a) Get moved around, sometimes far outside their expertise.
	   ("Can you explain RISC CPUs to me? I've just
	   moved over from the Food Column and I have to get up to speed
	   on this high-tech stuff.")
	b) Are faced with absorbing huge amounts of contradictory data/
	  marketeering from contentious companies, with buzzwords that
	  change quickly.
	c) Must often choose between quotable, but shallow "quick hits"
	  versus really hard work to not only figure out complex issues but
	  then make them simple enough to understand.
	d) Have ugly deadlines and few column-inches.

4) Then, there are plenty who have specific axes to grind, do not even
attempt to verify facts, and know very little about which they write.

(Personally, I think most press people fit category (3); I have sympathy
for anybody who makes an honest effort to sort things out; although it wasn't
FOOD, I've seen 3a in action more than once.)

-----
I don't know specifically about this case mentioned above; however, I have
more than once heard / given press interviews that went like:

reporter: we understand SGI is in the movie business and ships dinosaurs,
and this business is under attack by PCs.

SGI: entertainment industries are 15-20%; we don't ship dinosaurs, although
our products help our customers do that; they win awards, and we're proud
that we can help.  Entertainment is a very important market to us, and we're
doing just fine with it... it's 20% of our business...
30% is in manufacturing industries, where many products you
use, or your life depends on, were at least partly designed with SGIs.
X% is in large data management - 6 of the largest 15 UNIX-based data
warehouses are SGIs.   Y% is in Web.
Z% is in chemistry and biology, many drugs are designed on SGIs ... etc,
etc, with many examples, dutifully written down.

resulting article headline:
	"SGI's core dinosaur business under attack by PCs." 

(Well, not quite ... but I've seen things very close, where I *knew* the
discussion from which the article was derived.  The plain fact is that
many reporters have seen the movies, and have readers who have, and they
all relate to that, so they use it.)


-- 
-john mashey    DISCLAIMER: <generic disclaimer, I speak for me only, etc>
UUCP:    mash@sgi.com 
DDD:    415-933-3090	FAX: 415-967-8496
USPS:   Silicon Graphics/Cray Research 6L-005, 2011 N. Shoreline Blvd, Mountain View, CA 94039-7311

Subject: Re: 1980s the "golden age" of super-computing
From: mash@mash.engr.sgi.com (John R. Mashey)
Date: Nov 20 1996
Newsgroups: comp.sys.super,comp.arch

(Note: this doesn't have much to do with comp.arch, except insofar as it
contains cautionary tales on evaluating information, at least some oif
which is related to computer architecture):

In article <3292C82C.47B2CC51@clarkcom.com>, Glen Clark
<glen@clarkcom.com> writes:

|> But in fairness to BW... and acknowledging that there are some spins
|> SGI management might wish to change... I didn't think the article
|> was SGI-hostile. On the contrary, I thought they gave an unhurried
|> and involved analysis which gave no indication to me that somebody 
|> was trying to make it fit in X column inches. While saying that SGI
|> was having less success than it wished at creating a public image
|> that they sell something besides dinosaurs, I thought it was, in
|> general, pro Cray/SGI.

As I noted, I haven't seen the article ... and didn't imply it was
anti-SGI [in fact, sometimes I've read very pro-SGI articles that made me
cringe, because they in fact presented a distorted view].  What's sometimes
weird is the following sequence:
(a) Tell reporters about A+B+C+D, article focuses on C.
(b) A few years later, reporter from same magazine says "You're having
trouble shaking the image of C."  It may well be true, but it's ironic.
Like the stock market, the press sometimes gets into "lemming mode",
or even "recursive lemming mode" where what's printed becomes
increasingly divorced from any reality.  One runs across multiple
articles that only:
	(a) Reword press releases.
	(b) Quote other publications.
	(c) Quote analysts.
	(d) Quote customers/users of relatively low impact (because the
	big ones usually won't be quoted).
and basically have no new information content or analysis.  This is not
a criticism, just an observation to promote healthy skepticism.

|> BTW. Just got back from SC'96 about an hour ago. The two PhD candidates
|> I drove with have christened the O2 "the toaster" (which I guarantee 
|> will stick irrevocably) and have vowed to get a pair because it's
|> "cute".

Hopefully there are better reasons ... I want one because it will go
with my Espressigo (a purple Indigo turned into an espresso machine) at home
in the kitchen.

-- 
-john mashey    DISCLAIMER: <generic disclaimer, I speak for me only, etc>
UUCP:    mash@sgi.com 
DDD:    415-933-3090	FAX: 415-967-8496
USPS:   Silicon Graphics/Cray Research 6L-005, 2011 N. Shoreline Blvd, Mountain View, CA 94039-7311