From chemistry-request@www.ccl.net Fri Mar 5 01:43:10 1999 Date: Fri, 5 Mar 1999 01:42:04 -0500 From: Jean Debord Subject: Re: Biological relevance of aluminium Sender: Jean Debord To: CCL Dear Colleagues, I have forwarded the questions about the biological importance of aluminium to my friend and colleague Jean-Claude Bollinger, who gave some additional references: * P Zatta & al., "Al toxicity, the relevant role of metal speciation", Analusis 26, M72 (1998) (a paper copy is available on request to JCB, please send your postal address!) * JR Duffield & al., "low molecular mass Al complex speciation in biofluids", J. Coord. Chem. 23,277 (1991) * H Jacqmin & al., "components of drinking water & risk of cognitive impairment in the elderly", Am. J. Epidemiol. 139, 48 (1994) * CR Harrington & al., "Alzheimer's-disease-like changes in tau protein processing: association with Al accumulation in brains of renal dialysis patients", Lancet 343,993 (1994) Dr Bollinger insists on the importance of carrying out a speciation analysis of aluminium, either experimentally or theoretically. I hope you will find these references useful. Best regards, Jean Debord Limoges, France From chemistry-request@www.ccl.net Fri Mar 5 07:55:03 1999 From: Jochen Kuepper Subject: Re: AMD K6 DGEMM and L-3 BLAS Date: Fri, 5 Mar 1999 13:53:02 +0100 To: chemistry@www.ccl.net On Don, 04 Mär 1999 Matt Challacombe wrote: >To my knowlege, there is no optimized BLAS for the AMD. >If you have the time, your best bet is probably to use >PHiPAC to optimize a DGEMM specifically for your chip >(http://www.icsi.berkeley.edu/~bilmes/phipac) or ATLAS >(www.cs.utk.edu/~rwhaley/ATLAS/errata.html). My guess >is that these options will result in a faster DGEMM >than simply using the Greg Henry Intel optimized version. > >You can then obtain DGEMM based level 3 BLAS routines >from netlib, which will take advantage of your optimized >DGEMM. Unless you are doing exclusively level 1-2 ops, >having an optimized level 3 BLAS should make a big >difference. Absolutely. >If you (or anyone else in CCL land) does the optimization >(it can take 1cpu week+), please let me know how it went. Matt, thanks for that input. Looks really nice, I don't have the time to do that. So maybe someone else can go through the stuff and generate one. Or isn't there any interest out here ? Greetings, Jochen ----------------------------------------------------------------------- Jochen Küpper Heinrich-Heine-Universität Düsseldorf jochen@uni-duesseldorf.de Institut für Physikalische Chemie I Universitätsstr. 1, Geb 26.43 Raum 02.29 phone ++49-211-8113681 40225 Düsseldorf fax ++49-211-8115195 Germany http://www-public.rz.uni-duesseldorf.de/~jochen ----------------------------------------------------------------------- From chemistry-request@www.ccl.net Fri Mar 5 04:19:00 1999 Received: from bacchus.pc1.uni-duesseldorf.de (root@bacchus.pc1.uni-duesseldorf.de [134.99.152.11]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id EAA08606 Fri, 5 Mar 1999 04:18:55 -0500 (EST) Received: from bacchus.pc1.uni-duesseldorf.de (jochen@localhost [127.0.0.1]) by bacchus.pc1.uni-duesseldorf.de (8.8.7/8.8.7) with SMTP id KAA21372; Fri, 5 Mar 1999 10:18:32 +0100 From: Jochen Kuepper Organization: Heinrich-Heine-Universitaet Subject: Re: CCL:Dore metall Date: Mon, 11 Apr 2135 16:36:39 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain To: CHEMISTRY@www.ccl.net References: MIME-Version: 1.0 Message-Id: <99030510183200.21336@bacchus.pc1.uni-duesseldorf.de> Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by www.ccl.net id EAA08607 Status: RO Content-Length: 1624 On Mit, 03 Mär 1999 Richard Wood wrote: >This is not intended to be snotty or anything like that. But, doesn't >anybody know how to use a library anymore? Excuse me, but I have comment on that as well: Looking at this list it looks more & more like that. But there are really good alternatives: Go into your library and ask the librarian what possibilities you have to search for literature ! There's Chemical Abstracts, to tell you just one possibility, and there are many others with different purpose, differnt topics to cover, ... . Also you can search the web, look at online Journals indices, use online databases to search for the needed literature. Actually for me the most effecient think to do is to go to our service librarian. She does perform a online search together with me, so we really get that stuff done within 20 to 30 min. That's because I know what I want to know and she knows where/how to search for it. Please think about that possibilities before you ask this list for literature on topic x or y, next time. I would like to see some more discussion on content here and a lot less literature. Greetings, Jochen ----------------------------------------------------------------------- Jochen Küpper Heinrich-Heine-Universität Düsseldorf jochen@uni-duesseldorf.de Institut für Physikalische Chemie I Universitätsstr. 1, Geb 26.43 Raum 02.29 phone ++49-211-8113681 40225 Düsseldorf fax ++49-211-8115195 Germany http://www-public.rz.uni-duesseldorf.de/~jochen ----------------------------------------------------------------------- From chemistry-request@www.ccl.net Fri Mar 5 08:03:50 1999 Received: from Mercury.unix.acs.cc.unt.edu (mercury.acs.unt.edu [129.120.220.1]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id IAA09315 Fri, 5 Mar 1999 08:03:49 -0500 (EST) Received: from jove.acs.unt.edu (6930@jove.acs.unt.edu [129.120.220.41]) by Mercury.unix.acs.cc.unt.edu (8.8.8/8.8.8) with ESMTP id HAA05854; Fri, 5 Mar 1999 07:02:50 -0600 (CST) Received: from localhost (aiz0001@localhost) by jove.acs.unt.edu (8.8.8/8.8.8) with SMTP id HAA01153; Fri, 5 Mar 1999 07:02:47 -0600 (CST) Date: Fri, 5 Mar 1999 07:02:47 -0600 (CST) From: Anita Ilze Zvaigzne Reply-To: Anita Ilze Zvaigzne To: CHEMISTRY@www.ccl.net Subject: Re: CCL:Dore metall In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Please do not take this message in the wrong spirit. It is intended as a gentle reminder that not everyone on this list may have access to good library facilities or chemical literature databases. Thus, some researchers may need to ask other colleagues in the international chemistry community for assistance. I certainly am indebted to every member of this list who has summarized responses, suggested references or answered questions posted on this list. There is much to learn from these discussions. Thank you. Sincerely, Anita Zvaigzne ================================================================= On Thu, 4 Mar 1999, Richard Wood wrote: > This is not intended to be snotty or anything like that. But, doesn't > anybody know how to use a library anymore? > > Richard > > On Wed, 3 Mar 1999, Milan Antonijevic wrote: > From chemistry-request@www.ccl.net Fri Mar 5 12:14:39 1999 Received: from magnum.cooper.edu (magnum.cooper.edu [199.98.16.4]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with SMTP id MAA20022 Fri, 5 Mar 1999 12:14:38 -0500 (EST) Received: from robin.cooper.edu by magnum.cooper.edu with SMTP id AA04765 (5.65c/IDA-1.4.4 for ); Fri, 5 Mar 1999 12:13:20 -0500 Received: by robin.cooper.edu (SMI-8.6/SMI-SVR4) id MAA07615; Fri, 5 Mar 1999 12:13:27 -0500 Message-Id: <199903051713.MAA07615@robin.cooper.edu> Subject: Re: CCL:QC journals 'in trouble' Date: Fri, 5 Mar 1999 12:13:27 -0500 (EST) From: "Prof. Robert Topper" To: chemistry@www.ccl.net, Patrick.Bultinck@rug.ac.be In-Reply-To: <199903051327.IAA05043@bulldog.unca.edu> from "Jaime Martell" at Mar 5, 99 08:28:23 am X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII Hello all, > I have a hard time believing that a publication in JACS will not > score you enough points. Regarding physical chemistry journals, why > not publish in J. Phys. Chem.? In the ACS series, it is second in > reputation only to JACS, and surely must have a high impact factor. > Between J. Phys. Chem. and J. Chem. Phys. you will find most of the > highest quality theoretical chemistry papers. > > Sorry to disagree so strongly, but it seems to me that it's slightly unrealistic to expect that all articles touching on quantum chemistry should from now on only be published in one of three journals (JACS, JCP and JPC). I don't even see why that should be a good thing. Also, I have found that there are high-quality articles published in Chemical Physics Letters, Theoretical Chemistry Accounts, Computer Physics Communications, International Journal of Quantum Chemistry, Internet Journal of Chemistry, Journal of Comp. Chem., as well as THEOCHEM (I personally have seen some good articles in THEOCHEM, and I myself have published there, as have some fairly prominent scientists). I have also seen some articles in JCP, JACS and JPC that I think are ABYSMAL and should never have been published there. As far as JCP is concerned, I know some very prominent scientists that have sworn to me that they will never publish there again. If they publish elsewhere, do you think that their work is not worth reading? I'm sure you would still be willing to read an article by Y.T. Lee in Chemical Physics Letters... My personal opinion is that all of this bibliometric measurement of impact is totally useless. Some articles will have short-term impact but no long-term importance; some will be ignored when first published but later be found to be important; some will be classics forever; some will properly rest on the ash-heap of the history of science. None of the measures can tell you whether a journal is worth subscribing to, because none of them can tell whether a particular article in a particular journal is, or will ever be, useful to the scientists and engineers in a particular institution. What most scientists really need is complete and total access to all of the peer-reviewed literature. The only reason that this doesn't happen is....money. And because librarians don't have enough money, they often resort to the use of bibliometric schemes in order to help them make tough decisions about what to keep and what not to keep. I have nothing but sympathy for them, but "impact factors" are virtually useless measures of a journal's worthinesss for inclusion in an institution's library. Journal subscription prices in the sciences are outrageously expensive. If you compare our costs to those of the nonscience journals, it's a crime how much libraries are being charged. So only the wealthy libraries can afford complete collections... and they are in effect subsidizing other libraries which borrow copies of journals they themselves can't afford to buy. I myself work at an institution which pays a neighboring university for access to its journals, so that we don't have to buy them ourselves. Between that access, the New York Public Library, and the occasional kindness of neighboring universities like Columbia, Rutgers and Seton Hall, I manage. But it definitely makes it harder to get work done when you are running all over the tri-state area just to get a peek at an article. In that sense, if there were only 3 places to pubish, I guess When will this madness end?! When authors and editors revolt against these ridiculous journal prices! And along those lines, I'd like to salute the Internet Journal of Chemistry-which is totally free, maintains high standards, is fully indexed, and serves the community of theoretical chemistry well (since it's online, no more mailing preprints!!). I also respect the efforts of the ACS and APS, which are both nonprofit organizations, to keep journal prices at reasonable and affordable levels. Hoping for a better future for our library resources, Robert From chemistry-request@www.ccl.net Fri Mar 5 12:14:58 1999 Date: Sat, 06 Mar 1999 01:16:30 +0800 From: Wing Lok Abe Kurtz Chiu To: chemistry@www.ccl.net Subject: Re: CCL:G:Summary: AMD-K6/linux for G98 Hi all, I think we must be careful during the comparison of the speed of CPU. Using different program as a benchmark to test certainly will get different result. In the Tom's Hardware Guide(http://www4.tomshardware.com/releases/99q1/990223/cpu-news-14.html), it states that "Floating point calculations have never been a friend of the K6 or K6-2, when using them for a comparison with Intel CPUs. 3DStudioMax needs pure FPU-power. Let us see how K6-3 and it’on-die L2-cache are doing here. Well, there isn't much of a difference between with or without the on-die L2-cache. 3DstudioMax is one of the programs that you don't really want to run on a K6-2 or K6-3, unless you've got quite a lot of time to waste." In the above site, you also can get the different CPU information. Regards, Kurtz Chiu NMR group, Dept. of Chemistry, CUHK Bernd Melchers wrote: > 'Jerry C.C. Chan wrote:' > > > > > > K6 is slower than PentiumII (with the same clock). Especially on floating > > point operations. This is not OS dependent, so I wouldn't expect better > > some words again to this theme: > there are benchmarks comparing the brand new AMD-K6-III (*not* the AMD-K6-2-3D!) > against the Pentium III which show that even the AMD-K6-III@400MHz outperforms the > Pentium-III@500MHz (for a special application benchmark!): > http://www.ga-source.com/cgi-bin/gaimage.cgi?images/amd/Slide6.JPG > > it is not only the floating point unit, which makes the Pentium-II and the Celeron > fast, it is also the builtin second level cache. The AMD-K6-III now has also > an on dye second level cache. The floating point unit itself is not faster > than the one of the AMD-K6-II. > > Some additional points: > Celeron CPUs are using 66 MHz Busspeed, compared to 100 MHz for Pentium II and K6. > Celeron and K6 CPUs are not able to do multiprocessing. > > Compilers usually optimize the code for Pentium or Pentium-II CPUs, but there is > a compiler which optimize also for the K6: pgcc, have a look at > ftp://ftp.goof.com/pub/pcg/source/ > > But Gaussian98 and CHARMM is not able to be compiled by the g77 part of > pgcc or egcs . > > Ciao > Bernd Melchers > > -- > Bernd Melchers | melchers@FU-Berlin.DE > Freie Universitaet Berlin | "We don't write software, we compose it." > AG Macromolecular Modelling - Prof. Dr. E.W. Knapp > for more information see http://userpage.chemie.fu-berlin.de/~melchers/ > > -------This is added Automatically by the Software-------- > -- Original Sender Envelope Address: chemistry-request@www.ccl.net > -- Original Sender From: Address: melchers@chemie.fu-berlin.de > CHEMISTRY@www.ccl.net: Everybody | CHEMISTRY-REQUEST@www.ccl.net: Coordinator > MAILSERV@www.ccl.net: HELP CHEMISTRY or HELP SEARCH | Gopher: www.ccl.net 73 > Anon. ftp: www.ccl.net | CHEMISTRY-SEARCH@www.ccl.net -- archive search > Web: http://www.ccl.net/chemistry.html From chemistry-request@www.ccl.net Fri Mar 5 13:46:25 1999 From: Jochen Kuepper To: chemistry@www.ccl.net Subject: Re: CCL:G:Summary: AMD-K6/linux for G98 Date: Fri, 5 Mar 1999 19:36:14 +0100 On Fre, 05 Mär 1999 Bernd Melchers wrote: >'Jerry C.C. Chan wrote:' >> >> >> K6 is slower than PentiumII (with the same clock). Especially on floating >> point operations. This is not OS dependent, so I wouldn't expect better > >some words again to this theme: >there are benchmarks comparing the brand new AMD-K6-III (*not* the AMD-K6-2-3D!) >against the Pentium III which show that even the AMD-K6-III@400MHz outperforms the >Pentium-III@500MHz (for a special application benchmark!): >http://www.ga-source.com/cgi-bin/gaimage.cgi?images/amd/Slide6.JPG > >it is not only the floating point unit, which makes the Pentium-II and the Celeron >fast, it is also the builtin second level cache. The AMD-K6-III now has also >an on dye second level cache. The floating point unit itself is not faster >than the one of the AMD-K6-II. Actually there was a good comparison at many different levels in the (excelent) German magazine c't; it covered many pages (~30, IIRC): - In standard application benchmarks K6-III and Pentium-III are very close, when compared at the same clock frequency or at AMD-450 against Pentium-500. Mostly the Pentium was very slightly ahead. - Looking at software that utilizes 3D, AMD got it's winners, sometimes with a good distance. - When looking at the assembler level (mostly aiming at graphics stuff), they got to the result: Exactly the same (if programmed well) ! Actually the Pentium III won when they used the new enhancement (ISSE ?), but you cannot expect software to use it until next year, probably :-( They came to the conclusion, that the iron basicallygives you the same, so there are to things left: pricing and the very good developer utilities Intel offers, AMD not. Still, I am just waiting to get the AMD-III at a good price and then will put one into my home-box :-) I am writing this lines from an AlphaStation though, that just beats all of them :-) Greetings, Jochen ----------------------------------------------------------------------- Jochen Küpper Heinrich-Heine-Universität Düsseldorf jochen@uni-duesseldorf.de Institut für Physikalische Chemie I Universitätsstr. 1, Geb 26.43 Raum 02.29 phone ++49-211-8113681 40225 Düsseldorf fax ++49-211-8115195 Germany http://www-public.rz.uni-duesseldorf.de/~jochen ----------------------------------------------------------------------- From chemistry-request@www.ccl.net Fri Mar 5 15:13:31 1999 From: lmclaugh@knowledgefoundation.com Date: Fri, 05 Mar 1999 15:10:04 -0500 To: chemistry@www.ccl.net Subject: Mesoscale Modeling Dear CCL'ers: We are organizing a meeting on Mesoscale Modeling for materials and chemicals (non-pharmaceutical). We'd like to provide actual examples of how organizations are fostering the integration of atomistic, molecular and engineered data. Our goal is to broaden understanding of this approach and further it toward widespread commercial implementation. We'd also like to get the perspective of those that are developing software tools for this purpose. As a side issue, we may also address Computational Modeling of Defects in Materials. Does anyone have suggestions for researchers that I might contact for more information? Any help is greatly appreciated. Best regards, Linda Linda McLaughlin The Knowledge Foundation, Inc. 101 Merrimac, Boston MA 02114 USA Tel (617) 367-7979, ext. 207 Fax (617) 367-7912 Email lmclaugh@knowledgefoundation.com www.knowledgefoundation.com From chemistry-request@www.ccl.net Fri Mar 5 16:46:13 1999 From: "Jaime Martell" Date: Fri, 5 Mar 1999 16:47:28 -500 Subject: Re: CCL:QC journals 'in trouble' Reply-to: jmartell@bulldog.unca.edu To: chemistry@www.ccl.net > > I have a hard time believing that a publication in JACS will not > > score you enough points. Regarding physical chemistry journals, why > > not publish in J. Phys. Chem.? In the ACS series, it is second in > > reputation only to JACS, and surely must have a high impact factor. > > Between J. Phys. Chem. and J. Chem. Phys. you will find most of the > > highest quality theoretical chemistry papers. > > > > > Sorry to disagree so strongly, but it seems to me that > it's slightly unrealistic to expect that > all articles touching on quantum chemistry should from now on > only be published in one of three journals (JACS, JCP and JPC). I > don't even see why that should be a good thing. > I agree that is unrealistic, and not desirable. I merely suggested it as a way to get around the requirement for publishing in high impact journals. > Also, I have found that there are high-quality articles published in > Chemical Physics Letters, Theoretical Chemistry Accounts, Computer > Physics Communications, International Journal of Quantum Chemistry, > Internet Journal of Chemistry, Journal of Comp. Chem., as well as > THEOCHEM (I personally have seen some good articles in THEOCHEM, and > I myself have published there, as have some fairly prominent > scientists). I have also seen some articles in JCP, JACS and JPC > that I think are ABYSMAL and should never have been published there. > As far as JCP is concerned, I know some very prominent scientists > that have sworn to me that they will never publish there again. If > they publish elsewhere, do you think that their work is not worth > reading? I'm sure you would still be willing to read an article by > Y.T. Lee in Chemical Physics Letters... > I should have mentioned some of these, esp. Chem.Phys. Lett., but we're starting to get into lower impact factors, the original posters problem. I agree there are abysmal articles in any journal, and good articles in any journal. I will read articles by Y.T. Lee and others of his eminent standing in Chemical Physics Letters or wherever they choose to publish. In my personal experience (admittedly not a scientific survey, but others have agreed with me) Theochem has a higher degree of lower quality articles than the other aforementioned journals, and it has a lower impact factor. Perhaps there is a correlation there. > My personal opinion is that all of this bibliometric > measurement of impact is totally useless. Some articles will have I'm inclined to agree it has limited value. > > Journal subscription prices in the sciences are outrageously > expensive. If you compare our costs to those of the nonscience > journals, it's a crime how much libraries are being charged. Yes. Int. J. Quant. Chem. priced themselves out of a lot of libraries several years ago. A good quality journal, but I don't get to read it. Sorry for wasting bandwidth on this. I wanted to clear up some misconceptions on my first reply. Any further replies will be to authors only. Jaime From chemistry-request@www.ccl.net Sat Mar 6 09:55:05 1999 Date: Sat, 06 Mar 1999 16:00:32 +0100 To: chemistry@www.ccl.net From: "Hr. Dr. S. Shapiro" Subject: GA for PLS Dear Colleagues; Does anyone know of a DOS or Win95 executable that uses a genetic algorithm for variable selection in least squares analyses? If so, kindly send details. Thanks, S. Shapiro toukie@zui.unizh.ch From chemistry-request@www.ccl.net Sun Feb 28 04:36:57 1999 Received: from nmr-v.ioc.ac.ru (root@nmr-v.ioc.ac.ru [193.233.3.213]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id EAA05312 Sun, 28 Feb 1999 04:36:41 -0500 (EST) Received: from cacr.ioc.ac.ru (val@localhost [127.0.0.1]) by nmr-v.ioc.ac.ru (8.8.7/8.8.7) with ESMTP id MAA01715 for ; Sun, 28 Feb 1999 12:50:33 +0300 Sender: val@nmr-v.ioc.ac.ru Message-ID: <36D91167.13FAD48F@cacr.ioc.ac.ru> Date: Sun, 28 Feb 1999 12:50:31 +0300 From: val Reply-To: val@cacr.ioc.ac.ru Organization: IOC X-Mailer: Mozilla 4.5 [en] (X11; I; Linux 2.0.30 i686) To: CHEMISTRY@www.ccl.net Subject: Re:CCL:Re:CCL:Summary: AMD-K6/linux for G98 > > For pinene using DFT/B3LYP with large basis sets, for > >example, the machine i bought for $11,000 was faster than a Cray T3E-600 > >with 4 processors and only 15% slower than a Cray T3E-1200 with 4 > >processors, both running parallel Gaussian 94. (The G94 calculations were > >done by Sosa and Frisch, so were probably pretty well optimized for the > >machine, see last July 15 number of J. Comput. Chem. Sosa et al. Vol 19, > >p. 1053). I haven't priced those machines but they must be >$100k. If you > >want to benchmark against one of your own machines, Pulay would be happy > >to do some calculations for you. > The comparison here is *completely* invalid. If you want to compare > the speed of machines, you must be running the same program. In this > case, a highly-optimized DFT implementation (the one from Pulay) is being > compared to an implementation which can charitably be called "not so > great", this does not say a single thing about the relative speeds of the > computers running those programs. > > Having Pulay do a benchmark for you would be altering two variables at the > same time: > (a) the hardware > (b) the software > it is impossible to decouple the effects of the two changes. > One more time, for good measure: > If you want reliable benchmarks of *hardware* performance, you must > compare the same programs on the different hardware. > Moreover, just comparing the same program is not enough. The program must be compiled with the same compiler. Even providing compilers >from one vendor were used, the compilers must produce assembler code with the same efficiency. Next, exactly the same operation systems should be used on the different platforms, isn't it? etc., etc., ... Ideal hardware benchmarking on a *real* ab-inito task seems to be a very difficult problem. I agree with Greg's opinion that 1) hardware benchmarking and 2) software benchmarking require different strategies in order to get correct results. However, I would not call the PII/Cray comparison completely invalid. This may be the third type of benchmarking: the best performance/price ratio. Another words, how to achieve the the best results within a given budget. For an average computational chemist this type of benchmarking is really important and the original post does give us useful information. Finely, I would suggest to support all kinds of benchmarks. Of course a care should be taken comparing the results from the different benchmark types. best regards, Valentin. ==================================================================== , , , , Valentin P. Ananikov |\\\\ ////| /////| NMR Group | \\\|/// | ///// | ND Zelinsky Institute of Organic Chemistry | |~~~| | |~~~| | Leninsky Prospect 47 | |===| | |===| | Moscow 117913 | | | | | | | Russia | | A | | | Z | | \ | | / | | / e-mail: val@cacr.ioc.ac.ru \|===|/ |===|/ http://nmr.ioc.ac.ru/Staff/AnanikovVP/ '---' '---' Fax +7 (095)1355328 Phone +7 (095)9383536 ==================================================================== From chemistry-request@www.ccl.net Sun Feb 28 11:19:54 1999 Received: from aragorn.ics.muni.cz (aragorn.ics.muni.cz [147.251.4.33]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id LAA06845 Sun, 28 Feb 1999 11:19:54 -0500 (EST) Received: from bilbo.chemi.muni.cz (bilbo.chemi.muni.cz [147.251.28.2]) by aragorn.ics.muni.cz (8.8.5/8.8.5) with ESMTP id RAA19392; Sun, 28 Feb 1999 17:19:53 +0100 (MET) Received: from x-ray.chemi.muni.cz (x-ray.chemi.muni.cz [147.251.28.33]) by bilbo.chemi.muni.cz (8.8.5/8.8.5) with SMTP id RAA07056; Sun, 28 Feb 1999 17:20:33 +0100 (MET) Message-Id: <199902281620.RAA07056@bilbo.chemi.muni.cz> Comments: Authenticated sender is From: "Jaromir MAREK" To: "Gregory A. Landrum" Date: Sun, 28 Feb 1999 17:19:53 +0000 CC: chemistry@www.ccl.net > Subject: CCL:Re: CCL:Summary: AMD-K6/linux for G98 ^^^^^^^ !! > > For pinene using DFT/B3LYP with large basis sets, for > >example, the machine i bought for $11,000 was faster than a Cray T3E-600 > >with 4 processors and only 15% slower than a Cray T3E-1200 with 4 > >processors, both running parallel Gaussian 94.... > > The comparison here is *completely* invalid. I must disagree with your oppinion, > If you want to compare the speed of machines, you must be running the same program. ^^^^^^^^^^^^^^^^^ because the speed of (DFT) codes on the dif. machines and not speed of machines was the subject of the original post. > Having Pulay do a benchmark for you would be altering two variables at the > same time: > (a) the hardware > (b) the software > it is impossible to decouple the effects of the two changes. ^^^^^^^^^^^^^^^^ No. We have problem (e.g. DFT/B3LYP optimization) and for instance two independent variables - speed and budget of money - for selecting of the best solution. Hardware (and software) platform itself IS NOT really important, especially if the results on all tested HW+SW platforms will be (almost) the same... Jaromir ******************************************************************************* Jaromir MAREK e-mail:marek@chemi.muni.cz tel. ++ 420-5-411-29-411 fax. ++ 420-5-412-11-214 X-ray Laboratory of Dept. of Inorganic Chemistry http://www.sci.muni.cz/~dusan/inorg.html & Laboratory of Biomolecular http://www.chemi.muni.cz/lbsd/lbsd.html Structure and Dynamics Faculty of Science http://www.sci.muni.cz/ Masaryk University http://www.muni.cz/ Kotlarska (street) 2 BRNO CZ 611 37 Czech Republic