Path: utzoo!utgpu!water!watmath!clyde!rutgers!ames!pioneer!eugene
From: eugene@pioneer.arpa (Eugene Miya N.)
Newsgroups: comp.arch
Subject: Re: Single tasking the wave of the future?
Message-ID: <3618@ames.arpa>
Date: 10 Dec 87 17:59:44 GMT
References: <201@PT.CS.CMU.EDU> <388@sdcjove.CAM.UNISYS.COM> <988@edge.UUCP> <1227@sugar.UUCP> <151@sdeggo.UUCP> <1423@cuuxb.ATT.COM> <439@xyzzy.UUCP> <440@xyzzy.UUCP> <36083@sun.uucp> <18@amelia.nas.nasa.gov>
Sender: usenet@ames.arpa
Reply-To: eugene@pioneer.UUCP (Eugene Miya N.)
Organization: NASA Ames Research Center, Moffett Field, Calif.
Lines: 41

In article <18@amelia.nas.nasa.gov> fouts@orville.nas.nasa.gov (Marty Fouts) writes:
>In article <36083@sun.uucp> ram@sun.UUCP (Renu Raman, Sun Microsystems) writes:
>>how long will uni-processor machines keep parallel processors at bay?
>
>Well, as Gene Amdahl would say, it's not that simple.  Parallel
>processing, at least in the sense of throwing multiple processors at a
>single problem is currently difficult to use.
 . . .
>Saturation of usable HW technology is close.

Merry Christmas, Renu:
Oh, the "how long small can you make a Cray?" or Dan Ingall's "Cray on a
wrist idea?

Well, 1) we have some parallel processors now.  The most effective ones
are in loosely coupled, lightly stressed situations.  Just look at IBMs,
Univacs, etc. we've had them for a long time (but distinct process (jobs)
for processors). I think the Sequent/Flex/et al way is the way to
currently build machines, not just workstations (not enough power), but
you have to make it cheaper.  Marty, notes all the problems of other
types.

2) Since Marty didn't answer the question, I would hazard a generic
guess (I don't know anything) of about 10 more years (far enough into
the future?).  There is a Rand Corp report by Sutherland, Meade, and
Everhart on the limitations of silicon (there are other similar reports)
which chronicles about 9 generations of Silicon.  They were "3rd gen"
back in 1976 when they wrote it, and we are sort of "5-6 gen" by their
comments.  After all, how fast can you push the 68K ;-)?  I would just
hate to have some life-saving application which needs just that little bit
more power that silicon could not provide.  If you believe in the
infinite power of computing, then you might be able to tell me when we
have a sequential 100 Teraflops machine.

All comments grossly over-generalized to answer this general question.
From the Rock of Ages Home for Retired Hackers:

--eugene miya, NASA Ames Research Center, eugene@ames-aurora.ARPA
  "You trust the `reply' command with all those different mailers out there?"
  "Send mail, avoid follow-ups.  If enough, I'll summarize."
  {uunet,hplabs,hao,ihnp4,decwrl,allegra,tektronix}!ames!aurora!eugene