From owner-chemistry@ccl.net Mon Oct 31 01:01:01 2005 From: "janl*speakeasy.net" To: CCL Subject: CCL: CCL unmoderator asks you to learn perl regular expressions Message-Id: <-29852-051031003530-13900-XS9ZZdKgX045BGT2mR9KbA-x-server.ccl.net> X-Original-From: janl * speakeasy.net Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="iso-8859-1" Date: Mon, 31 Oct 2005 05:35:25 +0000 MIME-Version: 1.0 Sent to CCL by: janl:_:speakeasy.net Dear CCL, To help you learn perl regular expressions (you will need it to set up your preferences/filters for CCL mail) I wrote a Perl regexp training form (use it... I really love it {:-)}) at: http://www.ccl.net/cgi-bin/ccl/regexp/test_re.pl and for those who want to learn regexp, I wrote a tutorial at: http://www.ccl.net/chemistry/resources/tips/regular_expressions.shtml Links are also provided on the CCL home page. Learn perl: it will make you more efficient, rich, and respected by your friends and enemies, and significant ones. As to when the CCL profiles will be available... Soon, is the best answer I can give. They need to be checked for bugs and possible side effects (e.g., security, overhead, etc...). Jan The CCL Unmoderator From owner-chemistry@ccl.net Mon Oct 31 03:11:00 2005 From: "janl;;speakeasy.net" To: CCL Subject: CCL: Few more things from CCL Unmoderator Message-Id: <-29853-051031030757-26940-rIiCCnFuDsqIIdIWZd4AHQ ~ server.ccl.net> X-Original-From: janl]^[speakeasy.net Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="iso-8859-1" Date: Mon, 31 Oct 2005 08:07:51 +0000 MIME-Version: 1.0 Sent to CCL by: janl-.-speakeasy.net Dear CCL, Before my face hits the keyboard: 1) Look at new Conference Page on CCL: http://server.ccl.net/chemistry/announcements/conferences/index.shtml Please support this effort. 2) I started to work on the Supporting CCL page and massive redo of "informational" pages. Check the first version of Supporting page at: http://server.ccl.net/chemistry/aboutccl/supporting/index.shtml Jan ccl]![ccl.net From owner-chemistry@ccl.net Mon Oct 31 06:54:00 2005 From: "Toomas Tamm tt-ccl%kky.ttu.ee" To: CCL Subject: CCL: GPL'ed quantum chemistry software Message-Id: <-29854-051031050931-30794-HTvq3ZcMnh1jKJmpIghU0Q|*|server.ccl.net> X-Original-From: Toomas Tamm Content-Disposition: inline Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Mon, 31 Oct 2005 11:10:16 +0200 Mime-Version: 1.0 Sent to CCL by: Toomas Tamm [tt-ccl%a%kky.ttu.ee] For information to anyone who might be interested in developing / using GPL'ed QC software: such projects do exist, one of them being MPQC, http://www.mpqc.org/ . Despite what its name implies (MPQC stands for Massively Parallel Quantum Chemistry), it also runs reasonably fast on single workstations and small clusters. Here is a list of current capabilities from their web-page: * Closed shell, unrestricted and general restricted open shell Hartree-Fock energies and gradients * Closed shell, unrestricted and general restricted open shell density functional theory energies and gradients * Second order open shell perturbation theory (OPT2[2]) and Z-averaged perturbation theory (ZAPT2) energies. * Second order closed shell Møller-Plesset perturbation theory energies and gradients. * Second order Møller-Plesset perturbation theory including an R12 correlation factor using an auxilary basis set [(ABS) MP2-R12]. Energies of closed-shell systems are supported. * Robust internal coordinate geometry optimizer that efficiently optimizes molecules with many degrees of freedom. Nearly arbitrary internal coordinate constraints can be handled. I am in no way associated with the development (I wish I had time for that... ) but I guess this may be turned into a major QC program if more people would contribute. On the solid-state / plane-wave side, a similar project is abinit ( http://www.abinit.org/ ) . -- Toomas Tamm e-mail: tt-ccl at kky.ttu.ee Chair of Inorganic Chemistry voice: INT+372-620-2810 Tallinn University of Technology fax: INT+372-620-2828 Ehitajate tee 5, EE-19086 Tallinn, Estonia http://www.kk.ttu.ee/toomas/ From owner-chemistry@ccl.net Mon Oct 31 13:38:00 2005 From: "Shobe, David dshobe!A!sud-chemieinc.com" To: CCL Subject: CCL:G: Gibb's free energy Message-Id: <-29855-051031115954-28897-h7v+U/Z01R6x9TQ7FULQkQ**server.ccl.net> X-Original-From: "Shobe, David" Content-class: urn:content-classes:message Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C5DE39.BD53FCF2" Date: Mon, 31 Oct 2005 17:40:01 +0100 MIME-Version: 1.0 Sent to CCL by: "Shobe, David" [dshobe]|[sud-chemieinc.com] This is a multi-part message in MIME format. ------_=_NextPart_001_01C5DE39.BD53FCF2 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable It's easy if you use one of the multi-step methods (G3, CBS-Q, etc.): a Gib= bs free energy function (in hartrees) is part of the final output. =20 Otherwise, thermodynamic data can be obtained from a freq calculation; the = entropy and the thermal energy are in the section after the frequencies tab= le.=20=20 =20 --David Shobe, Ph.D., M.L.S.=20 S=FCd-Chemie, Inc.=20 phone (502) 634-7409=20 fax (502) 634-7724=20 Don't bother flaming me: I'm behind a firewall.=20 ________________________________ > From: owner-chemistry:-:ccl.net [mailto:owner-chemistry:-:ccl.net]=20 Sent: Sunday, October 30, 2005 6:01 AM To: Shobe, David Subject: CCL:G: Gibb's free energy Hi, How to calculate the Gibb's free energy for a molecule using gaussian softw= are ? Sincerely, ________________________________ Appel audio GRATUIT partout dans le monde avec le nouveau Yahoo! Messenger T=E9l=E9chargez le ici ! =20=20 This e-mail message may contain confidential and / or privileged informatio= n. If you are not an addressee or otherwise authorized to receive this mess= age, you should not use, copy, disclose or take any action based on this e-= mail or any information contained in the message. If you have received this= material in error, please advise the sender immediately by reply e-mail an= d delete this message.=20 Thank you. ------_=_NextPart_001_01C5DE39.BD53FCF2 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
It's easy if you use one of the multi-step methods= (G3,=20 CBS-Q, etc.): a Gibbs free energy function (in hartrees) is part of the fin= al=20 output.
 
Otherwise, thermodynamic data can be obtained from= a freq=20 calculation; the entropy and the thermal energy are in the section after th= e=20 frequencies table. 
 

--David Shobe, Ph.D., M.L.S.
S=FCd-Chemie, Inc.
phone=20 (502) 634-7409
fax (502) 634-7724

Don't bother flaming me: I'm behind a=20 firewall.


From: owner-chemistry:-:ccl.net=20 [mailto:owner-chemistry:-:ccl.net]
Sent: Sunday, October 30, 2005 = 6:01=20 AM
To: Shobe, David
Subject: CCL:G: Gibb's free=20 energy

Hi,

How to calculate the Gibb's free energy for a molecule using gaussian=20 software ?

Sincerely,

Appel audio GRATUIT partout dans le monde avec=20 le nouveau Yahoo! Messenger
T=E9l=E9chargez=20 le ici !=20
This e-mail message may contain confidential and / or privileged infor= mation. If you are not an addressee or otherwise authorized to receive this= message, you should not use, copy, disclose or take any action based on th= is e-mail or any information contained in the message. If you have received= this material in error, please advise the sender immediately by reply e-ma= il and delete this message.
Thank you.
------_=_NextPart_001_01C5DE39.BD53FCF2-- From owner-chemistry@ccl.net Mon Oct 31 15:09:00 2005 From: "Shobe, David dshobe.:.sud-chemieinc.com" To: CCL Subject: CCL: help needed Message-Id: <-29856-051031115957-28918-9JUtfnUunPL3G2OzvntyQw]_[server.ccl.net> X-Original-From: "Shobe, David" Content-class: urn:content-classes:message Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset="iso-8859-1" Date: Mon, 31 Oct 2005 17:16:31 +0100 MIME-Version: 1.0 Sent to CCL by: "Shobe, David" [dshobe,sud-chemieinc.com] Sent to CCL by: "Perry E. Metzger" [perry-,-piermont.com] > By the way, the word is PROGRAMS, not CODES. "Code" can be used in certain > idioms, but it is a mass noun or a verb. It is never correct English to > pluralize a mass noun. I realize many computational chemists do that, but > it is not correct grammar. Do not do it. What you should have said is that "codes" (or using "code" as a count noun) is not idiomatic in the programming community. Remember... Always be careful disparaging even friends' grammatical habits, if just keeping linguistic matters neutral offers peace & quiet. :-) --David Shobe, Ph.D., M.L.S. Süd-Chemie, Inc. phone (502) 634-7409 fax (502) 634-7724 Don't bother flaming me: I'm behind a firewall. -----Original Message----- > From: owner-chemistry#,#ccl.net [mailto:owner-chemistry#,#ccl.net] Sent: Saturday, October 29, 2005 11:20 AM To: Shobe, David Subject: CCL: help needed "Alex. A. Granovsky gran++classic.chem.msu.su" writes: >> About a year ago I conducted a demonstration in which a single >> processor machine running NetBSD (not Linux, but the principle would >> apply to Linux as well) successfully rebuilt a large software system >> many times faster than a four processor Windows server. The Windows >> machine had more memory, and each individual processor was faster >> than the NetBSD machine's one processor. The reason? The Windows >> machine was unable to keep enough pages in memory to be able to keep >> its four processors running at 100%. The NetBSD box more or less put >> everything it needed into memory once and barely touched the disk >> again, so its processor hit 100% and stayed pinned there. If the >> Windows machine had been able to this, it would have easily >> outperformed the NetBSD machine, but since it could not, most of its >> four expensive processors were sitting idle most of the time. > > This is unfair comparison. In fact, we both know this, You perhaps do, but I don't. The comparison seems perfectly fair. The architecture of Windows makes it handle file caching in an extremely inefficient manner. This simply demonstrated it. > but it is much more unfair to give such an incorrect examples with > respect to people who do not. There is nothing even remotely incorrect about the example. > First of all, describing any benchmark, you must avoid any qualitative > statements. You must fully specify hardware configurations, OS > versions and relevant settings, any nonstandard drivers used, etc..., etc... It was a four processor machine running Windows 2003, with 4G of memory -- I can't remember the specific Xeon model being used or the speed, but each of the four were substantially faster than the 1.6Ghz processor in my laptop which none the less beat the heck out of the Windows box even though it had four times the memory and four processors. The Windows box also had much faster disks, and had drivers written by the manufacturers of the various pieces of hardware, unlike the machine that beat it up, which had drivers written by hobbyists in their spare time without original manufacturer specs. > You must also provide enough details on the particular benchmark used. Both the Windows machine and the NetBSD box were compiling about 4000 files of Java code, using the identical Sun supplied Java SDK, both compilations driven by Ant, using identical files, identical versions of Ant, etc. The performance of the compilers themselves was not at issue, though, because the reason the NetBSD box beat the Windows box wasn't that the NetBSD box had "faster compilers" or some such, but because the Windows box was running an OS with totally a worthless VM subsystem incapable of doing proper file caching. If the Windows machine had been running its processors flat out, it would have beaten the hell out of the NetBSD machine, but since it could not provide data to the processors fast enough for them to remain even at 10% of capacity, they were unable to finish the task as quickly. > In your case, you should specify compilers used on NetBSD and Windows > system, indicate compiler options of significant impact on the > compiler performance, None of that had the least impact. Although the systems were compiling absolutely identical codebases (both compilations were driven from identical versions of Ant as well), CPU performance was NOT a factor, code quality of the compilers was NOT a factor. The only factor was that the Windows VM subsystem sucks dingo kidneys, and thus no matter how much money one spent on the processors, the speed of the Windows system was completely bound on disk read performance, whereas the NetBSD box quickly cached all the data, so it was at 100% CPU at all times. > Second, we all know that there are many situations when running I/O or > memory-limited tests on SMP system in parallel results in worse > performance than using only single thread or single process. As it turns out, the Windows machine had essentially identical performance when running with only one processor. That's because it was I/O bound because of the exceptionally bad design of the Windows VM subsystem. > Next, if you used Cygwin to compile on Windows system, No, I didn't. > Finally, you must admit that this is not the type of benchmark of > interest/importance for typical computational chemist. I disagree. Many CC problems involve far more I/O than you would think IF THE OPERATING SYSTEM IS BADLY DESIGNED. Certainly not *all* programs behave this way, but many, in fact, are badly impacted by it. There are, as I said, also many other flaws in Windows. For example, network performance is badly impacted by the architecture of the Windows NDIS and TCP/IP layers. If you are running a tightly coupled compute cluster, that will hurt you. There are many other flaws, too. These are just a couple of examples. On top of that, there is the fact that (and I realize this is subjective) Windows is simply unmanageable. There is too much state stored in too many places. If you want to run 100 or 500 or 1000 Windows machines in a compute cluster you can devote your life to keeping the machines running and operating smoothly and not get anywhere, whereas with a bit of proper scripting you can make a cluster of Linux or other Unix systems sing while spending your real effort on your research and not on systems management. I've seen this hurt very hard in the real world. > In CC, building the project is not the most time-consuming step, So what? If you have to do lots of unneeded I/O with a data set that will not conveniently fit in buffer cache, you'll have the same problem. The compilation example was just an exemplar. The truth is, you hit exactly the same sorts of problems in large cluster situations all the time. The compiler example is something I'm quoting purely because it was dramatic, but you can get similar issues with all sorts of other kinds of tasks, including computational chemistry. > It seems you are not very familiar with typical Quantum Chemistry > codes, If you wish to believe that. By the way, the word is PROGRAMS, not CODES. "Code" can be used in certain idioms, but it is a mass noun or a verb. It is never correct English to pluralize a mass noun. I realize many computational chemists do that, but it is not correct grammar. Do not do it. >> I've personally conducted extensive benchmarking on this specific >> topic, and I've read enormous amounts of Microsoft documentation. > > So did I too. Well, then people will have to decide if they think you know what you're doing or if they think I do. Perryhttp://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txtThis e-mail message may contain confidential and / or privileged information. If you are not an addressee or otherwise authorized to receive this message, you should not use, copy, disclose or take any action based on this e-mail or any information contained in the message. If you have received this material in error, please advise the sender immediately by reply e-mail and delete this message. Thank you.