Path: utzoo!mnetor!uunet!yale!cmcl2!arizona!lm
From: lm@arizona.edu (Larry McVoy)
Newsgroups: comp.arch
Subject: Chen's claim
Message-ID: <4449@megaron.arizona.edu>
Date: 23 Mar 88 00:03:04 GMT
Reply-To: lm@megaron.arizona.edu (Larry McVoy)
Organization: U of Arizona CS Dept, Tucson
Lines: 19

Today I read an article (in Time magazine, I'll admit) that quoted Chen
as saying "Five years from now we should be at 100 billion gigaflops."
Let's see:  I know of 10 gigaflop machines so I suppose we probably
have 100 gigaflop machines around as well.  So he is saying that we'll
see 1,000,000,000 times improvement over the next five years.  Now, the
context was most definitely multi processor computing so that helps a
bit, but what do you think?  Do you think we'll see that kind of jump?
I don't know, I suppose anything is possible....  If he is right and
he does it, Cray blew it.

Ahh, here's some help:  he says he wants to problems that take 3 months
now in a day.  But that is only a 90x improvement.  If we had a
billion times improvement, we could do, in a day, a problem that would take
27,777 centuries (assuming I did my math correct, but it's around there)
of current computation time.  So I suspect that Time quoted him wrong.
Does anyone closer to Chen have any idea what was said/meant?
-- 

Larry McVoy	lm@arizona.edu or ...!{uwvax,sun}!arizona.edu!lm