RE: PAPER: Levels of Organization in General Intelligence

From: Ben Houston (ben@exocortex.org)
Date: Mon Apr 08 2002 - 07:13:49 MDT


> Oh, I've heard about GRAPE, but so far from astrophysics papers only.
> Mparticle MD is a special case for pharma, though, they're largely
> concerned with virtual screening of small molecule libraries, though some
> do peptides, of course.
>
> Ben, do you know any pharma company which uses clustered MD-GRAPE on a
> nontrivial scale (few 100 nodes), or are you bound by NDA?

Cool website you have there. I didn't know that this was your area of speciality. :-) Do you currently work for "molecular devices"?

Okay, well I got my information from the head of a research group at Novartis but I don't think it was too secret. I think that he was referring to the work described in these papers:

[1] Narumi T, Kawai A, Koishi T 2001. "An 8.61 Tflop/s Molecular Dynamics Simulation for NaCl with a Special-Purpose Computer: MDM" Proc. Supercomp. 2001. http://www.sc2001.org/papers/pap.pap205.pdf

[2] Narumi T, Susukita R, Ebisuzaki T, McNiven G, Elmegreen B. 1999. "Molecular dynamics machine:Special-purpose computer for molecular dynamics simulations", MolecularSimulation, vol. 21, pp.
401-415.

[3] Narumi T, et al. 1997. "Molecular Simulations with the MD-Grape Special-Purpose Computer" Proc. Int. Symp. On Supercomp.

[4] Komeiji Y, Yokoyama H, Uebayasi M, Taiji M, Fukushige T, Sugimoto D, Takata R, Shimizu A, Itsukashi K. 1996. "A high performance system for molecular dynamics simulation of biomolecules using a special-purpose computer." Pac Symp Biocomput 1996:472-87

-ben
www.exocortex.org

> -----Original Message-----
> From: owner-sl4@sysopmind.com [mailto:owner-sl4@sysopmind.com] On Behalf
> Of Eugen Leitl
> Sent: Monday, April 08, 2002 7:09 AM
> To: sl4@sysopmind.com
> Subject: RE: PAPER: Levels of Organization in General Intelligence
>
> On Mon, 8 Apr 2002, Ben Houston wrote:
>
> > Well for computer aided drug design where QM/MM is central there is an
> > add-in board for a high-end PC that can deliver 100 teraflops when used
> > in a 4 machine MMD cluster. Here's some info about it from IBM:
> > http://www.research.ibm.com/grape/
>
> Oh, I've heard about GRAPE, but so far from astrophysics papers only.
> Mparticle MD is a special case for pharma, though, they're largely
> concerned with virtual screening of small molecule libraries, though some
> do peptides, of course.
>
> Ben, do you know any pharma company which uses clustered MD-GRAPE on a
> nontrivial scale (few 100 nodes), or are you bound by NDA?



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:38 MDT