From: Peter C. McCluskey (pcm@rahul.net)
Date: Wed Nov 27 2002 - 10:44:49 MST
mez@apexnano.com (Ramez Naam) writes:
>From: Eugen Leitl [mailto:eugen@leitl.org]
>> > current modeling methods that are able to handle 10^9 atoms.
>I would love to see this reference. Since you mention GROMACS in your
>post, let me use that as an illustration of where I consider the state
>of the art to be. In Science 294 (2001), Bert L. de Groot and Helmut
>Grubmuller present a 100 nano-second simulation of 10^5 water
>molecules using the GROMACS code (which produces highly highly
>approximate solutions). At the time it was published, this article
>was hailed as one of the largest simulations ever done.
I suspect that the closest to what Gene is claiming is something like
the methods mentioned in:
http://www.hpc.csiro.au/External/NEC/1998Aug/quantum.htm
This apparently does quantum-level simulations of over 7.6 million atoms.
The shortage of comments from computational chemists about this breakthrough
suggests to me that this is a rather special-purpose solution.
That web page says "Dr Michalewicz's technique allows the study of
'imperfect' or complex solids". That probably means he started with
something that only worked on substances with a perfectly regular pattern
(i.e. something totally irrelevant to assembler design) and managed to
relax that restriction in some way that I haven't seen described. The
fact that he isn't describing what imperfections it can handle suggests
it isn't very general-purpose.
I suspect that the simulations that will be most relevant to early
assemblers will be even less accurate than GROMACS, and that a fair
amount of trial and error in the labs will be needed to come up with
something that works.
-- ------------------------------------------------------------------------------ Peter McCluskey | Free Jon Johansen! http://www.rahul.net/pcm |
This archive was generated by hypermail 2.1.5 : Wed Jan 15 2003 - 17:58:26 MST