Recent AMD K6-2 delivers up to 1.3 GFlops, and a peak memory bandwidth
of 0.8 GBytes/s. A 30 cm wafer in ~0.1 um technology made of evolvable
hardware is probably the end of the line for conventional
semiconductor photolitho. Similiarly, quantum dot cellular automata,
especially 3d-integrated are the end of the line for integration
density in conventional molecular circuitry.
If quantum computation works in solid state at high qubit numbers and
is sufficiently all-purpose (speeding up ab initio and cracking crypto
seem to be the more straightforward applications) we will have to reevaluate
the computation density anyway.
Whether we ever can harness anything beyond condensed phases for
computation purposes is open-ended.
ciao,
'gene
--59UH/NVKMw
Content-Type: message/rfc822
Content-Transfer-Encoding: 7bit
Return-Path: <majordom@cesdis1.gsfc.nasa.gov>
Received: from cesdis.gsfc.nasa.gov (cesdis1.gsfc.nasa.gov [128.183.38.12])
by liposome.genebee.msu.su (8.8.5/8.8.5) with SMTP id CAA29776
for <eugene@liposome.genebee.msu.su>; Fri, 26 Jun 1998 02:03:41 +0400
Received: by cesdis.gsfc.nasa.gov (4.1/SMI-4.1)
id AA16326; Thu, 25 Jun 98 17:12:19 EDT
Received: from alushta.NL.net by cesdis.gsfc.nasa.gov (4.1/SMI-4.1)
id AA16303; Thu, 25 Jun 98 17:09:16 EDT
Received: from moene by alushta.NL.net with UUCP id <6016-31064>; Thu, 25 Jun 1998 23:17:01 +0200
Received: by moene.indiv.nluug.nl (NX5.67e/NeXT-2.0)
id AA01843; Thu, 25 Jun 98 23:14:40 +0200
Message-Id: <9806252114.AA01843@moene.indiv.nluug.nl>
Content-Type: text/plain
Mime-Version: 1.0 (NeXT Mail 3.3 v118.2)
Received: by NeXT.Mailer (1.118.2)
References: <9806251530.AA24798@krang.parl.eng.clemson.edu>
Organization: Moene Computational Physics, Maartensdijk, The Netherlands
Precedence: bulk
From: Toon Moene <toon@moene.indiv.nluug.nl>
Sender: owner-beowulf@cesdis1.gsfc.nasa.gov
To: "Walter B. Ligon III" <walt@walt-pc.eng.clemson.edu>
Cc: beowulf@cesdis1.gsfc.nasa.gov
Subject: Re: Interim beowulf mailing list
Date: Thu, 25 Jun 98 23:14:33 +0200
> Of course, as I have said several times in the last few
> days, this really has nothing to do with the CESDIS
> Beowulf blackout. Yes, somewhere at NASA HQ there are
> some guys looking over it to see if it meets export
> requirements - but their main objective is to force the
> CESDIS guys back in line with "NASA policy" on
> distributing software. Other than that, nothing has
> changed in the last few weeks - all of this export control
> of Beowulf stuff is rumour and speculation. As of yet
> noone has ruled that Beowulf ideas, software, or
> components are export restricted.
Well, at least this makes sense; more so than the drivel you send
me personally.
Of course, NASA itself being silent on this issue made it very easy
for all sort of conspiracy theories being developed, and I
certainly felt being hit hard with this export restriction thing.
There's little doubt that quad-21264 motherboards exceed 2,000 MTOPs
and that means that my next PC *will* have export restrictions.
This is even independent of Beowulf stuff - although I have to use
either PVM or MPI on it to benefit from the four processors - as g77
doesn't do automatic parallellisation yet.
Regards,
-- Toon Moene (mailto:toon@moene.indiv.nluug.nl) Saturnushof 14, 3738 XG Maartensdijk, The Netherlands Phone: +31 346 214290; Fax: +31 346 214286 g77 Support: mailto:fortran@gnu.org; NWP: http://www.knmi.nl/hirlam--59UH/NVKMw--