From chemistry-request@server.ccl.net Sun Oct 17 11:47:34 1999 Received: from sunu450.rz.ruhr-uni-bochum.de (sunu450.rz.ruhr-uni-bochum.de [134.147.222.33]) by server.ccl.net (8.8.7/8.8.7) with SMTP id LAA10669 for ; Sun, 17 Oct 1999 11:47:33 -0400 From: Christoph.van.Wuellen@ruhr-uni-bochum.de Received: (qmail 27232 invoked from network); 17 Oct 1999 15:41:06 -0000 Received: from sgi249.rz.ruhr-uni-bochum.de (134.147.64.2) by mailhost.rz.ruhr-uni-bochum.de with SMTP; 17 Oct 1999 15:41:06 -0000 Received: (qmail 12214 invoked by uid 10283); 17 Oct 1999 15:40:46 -0000 Message-ID: <19991017154046.12213.qmail@sgi249.rz.ruhr-uni-bochum.de> Subject: DFT and dispersion forces To: chemistry@ccl.net Date: Sun, 17 Oct 1999 17:40:45 +0200 (MSZ) X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Just to get the discussion on formal grounds again: imagine two argon atoms at both sides of a "wall" (i.e., an infinite potential barrier). Since both densities do not overlap, the DFT exchange-correlation hole at a reference point at one side of the barrier cannot distinguish if at the other side there are MOVING electrons or a static charge distribution. Hence, while induction forces etc. are present in the DFT calculation, all dispersion interactions are necessarily absent (yes there is dispersion force between the two atoms). You will find the example esoteric, but two argon atoms at a distance where their densities practically do not overlap are just the same. Who can come up with a functional that gives dispersion in these cases: it is not impossible but necessarily non-local: if the reference point is on one side, the density at the other side must be taken into account as well. A gradient expansion ist not sufficient because of the convergence radius! ---------------------------+------------------------------------------------ Christoph van Wullen | Fon (University): +49 234 32 26485 Theoretical Chemistry | Fax (University): +49 234 32 14109 Ruhr-Universitaet | Fon/Fax (private): +49 234 33 22 75 D-44780 Bochum, Germany | eMail: Christoph.van.Wuellen@Ruhr-Uni-Bochum.de ---------------------------+------------------------------------------------ From chemistry-request@server.ccl.net Sun Oct 17 17:21:21 1999 Received: from mail.AlAkhawayn.ma (alakhawayn.ma [193.194.63.82]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id RAA11770 for ; Sun, 17 Oct 1999 17:21:15 -0400 Received: from location (fwinternet.alakhawayn.ma [193.194.63.129]) by mail.AlAkhawayn.ma (8.9.0/8.9.0) with SMTP id VAA16132 for ; Sun, 17 Oct 1999 21:17:39 GMT Message-Id: <199910172117.VAA16132@mail.AlAkhawayn.ma> X-Sender: J.Conceicao@mail X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0.2 Date: Sun, 17 Oct 1999 21:05:21 +0000 To: chemistry@ccl.net From: Jose Conceicao Subject: Computational Chemistry in Education Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Hi everyone Does anyone have any information/statistics on : 1. The number of universities/colleges currently using some form of computational modelling program in chemical education. 2. The number of universities/ colleges that teaches computational chemistry as part of their curriculum. Thanks in advance. Jo Dr. Jose J.B. Conceicao Assistant Professor Al Akhawayn University in Ifrane (AUI) School of Science & Engineering (SSE) B.P. 1904 Avenue Hassan II MOROCCO 53 000 Tel #: (212) 586-2143 Email:J.Conceicao@alakhawayn.ma http://www.alakhawayn.ma/~J.Conceicao/#top We the unwilling Led by the unqualified Have been doing the unbelievable so long with so little That we now attempt the impossible with nothing - Anonymous- From chemistry-request@server.ccl.net Sun Oct 17 12:31:57 1999 Received: from rzusuntk.unizh.ch (rzumail2.unizh.ch [130.60.128.10]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id MAA10871 for ; Sun, 17 Oct 1999 12:31:57 -0400 Received: [from rzusgi.unizh.ch (chburger@rzusgif.unizh.ch [130.60.68.19]) by rzusuntk.unizh.ch (8.8.5/SMI-5.31) with ESMTP id SAA29504; Sun, 17 Oct 1999 18:25:30 +0200 (MET DST)] From: "Dr. Peter Burger" Received: [(chburger@localhost) by rzusgi.unizh.ch (8.8.8/SMI-IRIX6.2/USAR_MAIL_V2) id SAA29722; Sun, 17 Oct 1999 18:25:30 +0200 (MDT)] Message-Id: <199910171625.SAA29722@rzusgi.unizh.ch> Subject: CCL: Re Beowulf/Chemistry/Athlon To: mthompson@seanet.com (Mark A. Thompson) Date: Sun, 17 Oct 1999 18:25:30 +0200 (MDT) Cc: CHEMISTRY@ccl.net In-Reply-To: <008001bf18b7$c8f47910$01808080@biggeek> from "Mark A. Thompson" at Oct 17, 99 08:53:40 am X-Mailer: ELM [version 2.4 PL25 PGP2] Content-Type: text Mark, for some info on how it's done check: http://www.cs.utk.edu/~rwhaley/ATL/INDEX.HTM Source & prebuilt libs are at: http://www.netlib.org/atlas/ !!! Alot of thanks go to Clint Whaley (the author) !!!! For the Athlon, you have to build one yourself or I can send you mine. Atlas just generates fast Blas-level 3 libs - you still need level 1 and 2 blas lib from other places. The Atlas generated dgemm on my PIII-550 runs at ca. 430 Mflops, which is way ahead (ca. 100 MFlop) of G. Henry's PentiumPro lib for Linux. On the Athlon-500 one can squeeze out approximately 540-550 MFlops. Assuming linear scaling, one can build for a reasonable amount of money (the 700 MHz Athlon is still too expensive) a 700 Mflop machine with the 650 MHZ version and still has the potential of > 1 GFlop once a dgemm routine is available that makes use of both FP (add, mul) pipelines. Peter ------------------------------------------------------------- Peter Burger Universitaet Zuerich chburger@aci.unizh.ch