From chemistry-request@ccl.net Fri Feb 14 00:13:00 1992 Date: Thu, 13 Feb 92 23:37:01 -0500 From: fredvc@esvax.dnet.dupont.com To: %chem@ccl.net Subject: OS-s versus Compilers Status: R A cautionary comment on this response. >>Regarding Doug Smith concerns on the use of VMS I would like >>to add that we have a group of VAX 8600 where programs >>run about 200 times slower than on cray ymp. We understood they >>should run 30-50 times slower considering the speed and vector >>capabilities of the ymp. We suspect it is a problem of VMS >>I would appreciate any comments on the conversion to ULTRIX (1) 8600's are not very fast VAXen. We found that speed doubled on going from a 8600 to an 8650, with more "juice" from an 8800 and a little more from a 6000-400. (2) CRAY Y-MP speed comes from several places: (a) a fast clock; (b) direct access to memory, rather than cache architecture; (c) very good SCALAR optimization; (d) Vector operations as an integral part of the system, rather as an add-on, along with good compiler recognition of vectorization opportunities. (3) The scalar/vector stuff (2c, 2d) is realized via the COMPILER, rather than the operating system, per se. There may be good reasons for converting from VMS to ULTRIX, but I would not place performance high on that list. Unless you can get a better compiler in the deal, I would expect things to be pretty much the same. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ FREDERIC A. VAN-CATLEDGE Scientific Computing Division || Office: (302) 695-1187 Central Research & Development Dept. || FAX: (302) 695-9658 The Du Pont company || P. O. Box 80320 || Internet: fredvc@esvax.dnet.dupont.com Wiilmington DE 19880-0320 || ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From chemistry-request@ccl.net Fri Feb 14 06:10:53 1992 Date: 14 Feb 92 11:10:29 EDT From: Subject: AM1 and charges in general To: chemistry@ccl.net Status: R E. M. EVLETH Dynamique des Interactions Moleculaires Universite Pierre et Marie Curie 4 Place Jussieu, Tour 22, Paris 75005 (1) 44 27 42 08 UDIM018 at FRORS31 Per the last request on the reliability of AM1 charges and other previous comments, the area is a can of worms. Reliability in relationship to what and by what methodology are they generated? First, there are no experimental atomic centered charges. Mulliken charges have their known faults but other methods exists. Many people are directed to modeling these charges from electrostatic potentials so that they can be used in various large scale modeling schemes and this electronic mail service abounds with comments, requests, and availability of softs which will do this. Mostly this is done from ab initio calculations but there are a number of papers in J. Comp. Chem. in the last 2 years on semiempirically based modeling and comparison with what a moderate ab initio basis set will give. BUT, as for what charges which are generated directly using AM1 and PM3 my experience is that they have little relationship to a Mulliken charges from a DZP level ab initio calculation. Likewise, there is little other than a qualitative relationship between ab initio electrostatic generated charges and either Mulliken charges and (in our limited experience) Bader type analyses (for highly ionic systems in the second row), about which there is a bit of current controversy (Perrin, JACS 1991). CONCLUSION, trust no one and nothing. One can argue that since has to use these charges in modeling systems, one should use the electrostatic (or electric field modeled) charges. But there are some methodological traps here too which I will not go into. From chemistry-request@ccl.net Fri Feb 14 11:04:34 1992 Date: Fri, 14 Feb 92 09:53:26 EST From: Kurt.Hillig@um.cc.umich.edu To: chemistry@ccl.net Status: R As a side note to the question of calculating atomic charges, there has been a lot of discussion in the literature of methods for structural prediction for van der Waals dimers. It's generally accepted that electrostatic interactions are a major contributor to the overall interaction. In general, methods based on atomic charges alone do not predict the structures or binding energies well. Furthermore, methods based on the distributed multipole model of Buckingham, Fowler and Stone appear to work fairly well only when one of the partners of the dimer is non-polar (e.g. in C2H4 - SO2), but usually fails when both components are polar (as in (SO2)2). What does it all mean? My best interpretation is that you need to be very careful when applying simple electrostatic models to intermolecular interactions. (Profound, isn't it?) So, I emphasize the cautions given about charge calculations - reliability is hard to define and harder to quantify (and, unfortunately, often impossible to achieve). From chemistry-request@ccl.net Fri Feb 14 19:08:43 1992 Date: Fri, 14 Feb 92 16:42:03 EST From: Shaun Black Subject: Access to software, archie servers To: chemistry@ccl.net Status: R Greetings all- I've noticed that quite a few requests for sites to FTP various software packages have appeared on CHEMISTRY in recent months. I thought I would mention an alternate route to locate such sites, namely the ARCHIE servers. Many of us may already know about this useful service, but for those who don't, I've detailed below the location of the ARCHIE servers and the basics of using them. ARCHIE is a shortened version of "ARCHIvE" and is a continuously updated database of about 900 FTP'able sites for innumerable software items and files. Telnet on to any of the (identical database) ARCHIE sites, as follows: quiche.cs.mcgill.ca 132.206.2.3 (The original; often slow and busy) archie.sura.net 128.167.254.179 (nic.sura.net) archie.rutgers.edu 128.6.18.15 (dorm.rutgers.edu) archie.ans.edu 147.225.1.2 (nis.ans.net) Login as ARCHIE, and no password is required. The two most useful commands are WHATIS (e.g., whatis protein) and PROG (e.g. prog ampac). WHATIS locates software by subject, and PROG locates software by FTP site. Ultimately you will need to use PROG to find one (usually of many) possible FTP addresses. To end type BYE, QUIT, etc. For international netters, ARCHIE is available at the following sites: archie.funet.fi 128.214.6.100 (Europe, not including the UK) archie.doc.ic.ac.uk 146.169.3.7 (puffin.doc.ic.ac.uk, UK) archie.au 128.184.1.4 (rana.cc.deakin.oz.au, Australia) I hope this information is useful to many of you. Cheers. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= - Shaun D. Black, PhD | Bitnet: black@ohstphrm.bitnet - = Ohio State University | Internet: black@ohstphrm.pharmacy.ohio-state.edu = - College of Pharmacy | Phone: (614) 292-3925 - = 500 West 12th Avenue | FAX: (614) 292-2435 = - Columbus, OH 43210-1291 | :-) (...Start every day with a smile...) - =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From jkl@ccl.net Mon Feb 24 23:10:28 1992 To: chemistry@ccl.net Subject: Address questions Date: Mon, 24 Feb 92 23:10:19 EST From: jkl@ccl.net Status: R If I may insert my twopence worth. There are frequent questions about e-mail/addresses of people on Computational Chemistry List. In many cases, I have this info, since they are (or were, for that reason...) on the list. I will gladly supply this info if asked directly, i.e., if you send mail to jkl@ccl.net or JKL@OHSTPY.BITNET. In my opinion it does not contradict the rule of the list which says that the whole list of subscribers is not publicly available. Otherwise close to 1000 people gets these inquires. Of course, the choice is yours. Just trying to help... Jan Labanowski jkl@ccl.net