From chemistry-request@ccl.net Mon Feb 7 22:08:59 2005 Received: from smtp.goldrush.com (smtp.goldrush.com [206.171.171.11]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j1838w5J000992 for ; Mon, 7 Feb 2005 22:08:58 -0500 Received: from Compaq (x2-04-0.goldrush.com [64.162.10.0]) by smtp.goldrush.com (8.12.8/8.12.8) with SMTP id j1838nS5019331; Mon, 7 Feb 2005 19:08:57 -0800 From: "Steve Bowlus" To: "Rodrigo Pontes" , Subject: RE: CPU usage Date: Mon, 7 Feb 2005 19:08:48 -0800 Message-ID: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0006_01C50D48.73DD0C40" X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) In-reply-to: <20050207172913.80630.qmail = = = web60103.mail.yahoo.com> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Importance: Normal X-MailScanner: Found to be clean X-MailScanner-SpamCheck: X-MailScanner-From: chezbowlus = = = goldrush.com X-Spam-Status: No, hits=3.6 required=7.5 tests=HTML_30_40,HTML_FONTCOLOR_BLUE, HTML_MESSAGE,MK_YAHOO_REDIR_01,MY_HTML_OBFU,WLS_URI_OPT_150 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------=_NextPart_000_0006_01C50D48.73DD0C40 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit I have observed the same behavior with another quantum code on a machine with the same CPU/OS combination. There was an earlier post to this list asking about what seemed to be a similar problem with the Xeon HT. I tentatively interpret the behavior that the job is given 100% of one of the HT execution streams, while the other stream is largely kept idle. I have not found a way to tune this behavior. If you look at ControlPanels >AdministrativeTools >ComputerManagement >Processors, you will see there are two CPUs shown. These can be uninstalled, but I haven't tried uninstalling one. Presumably, all CPU resources would be concentrated in a single execution stream, but I haven't found any documentation on this point. I also haven't found anything to suggest the second "CPU" can be _re_installed. I haven't talked to Intel or HP (who built my machine), either of whom could shed light on all these issues. Seems a pity that information on such a feature (or limitation) of HT technology would not be made readily available. sb -----Original Message----- From: Computational Chemistry List [mailto:chemistry-request = = = ccl.net]On Behalf Of Rodrigo Pontes Sent: Monday, February 07, 2005 9:29 AM To: chemistry = = = ccl.net Subject: CCL:CPU usage Dear Colleagues, We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz machine. We observed (in the Task manager / performance) that the CPU usage is only around 60%. Have someone noticed the same problem? How could us set the configuration to a 100% CPU usage? Thanks in advance, Rodrigo M. Pontes Departemanto de Quimica Universidade Estadual de Maringa Brazil ---------------------------------------------------------------------------- -- Yahoo! Acesso Gratis - Internet rapida e gratis. Instale o discador do Yahoo! agora. ------=_NextPart_000_0006_01C50D48.73DD0C40 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

I have=20 observed the same behavior with another quantum code on a machine with = the same=20 CPU/OS combination. There was an earlier post to this list asking = about=20 what seemed to be a similar problem with the Xeon HT. =

I=20 tentatively interpret the behavior that the job is given 100% of one of = the HT=20 execution streams, while the other stream is largely kept idle. I = have not=20 found a way to tune this behavior. If you look at ControlPanels=20 >AdministrativeTools >ComputerManagement >Processors, you = will see=20 there are two CPUs shown. These can be uninstalled, but I haven't = tried=20 uninstalling one. Presumably, all CPU resources would be = concentrated in a=20 single execution stream, but I haven't found any documentation on this=20 point. I also haven't found anything to suggest the second "CPU" = can be=20 _re_installed.

I=20 haven't talked to Intel or HP (who built my machine), either of whom = could shed=20 light on all these issues. Seems a pity that information on such a = feature=20 (or limitation) of HT technology would not be made readily=20 available.

-----Original = Message-----
From:=20 Computational Chemistry List [mailto:chemistry-request = = = ccl.net]On = Behalf Of=20 Rodrigo Pontes
Sent: Monday, February 07, 2005 9:29=20 AM
To: chemistry = = = ccl.net
Subject: CCL:CPU=20 usage

Dear Colleagues,

We are running Gaussian on a Windows XP/Pentiun-IV HT=20 3.2GHz machine. We observed (in the Task manager / performance) = that the=20 CPU usage is only around 60%. Have someone noticed the same = problem?=20 How could us set the configuration to a 100% CPU usage?

Thanks in advance,

Rodrigo M. Pontes

Departemanto de Quimica

Universidade Estadual de Maringa

Brazil

Yahoo!=20 Acesso Gratis - Internet rapida e gratis. Instale o discador do = Yahoo!=20 agora.

------=_NextPart_000_0006_01C50D48.73DD0C40-- From chemistry-request@ccl.net Tue Feb 8 10:12:16 2005 Received: from jaguar.mail.utk.edu (jaguar.mail.utk.edu [160.36.0.84]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18FCFi8025747 for < ccl.net>; Tue, 8 Feb 2005 10:12:16 -0500 Received: from [160.36.201.91] (herschel.chem.utk.edu [160.36.201.91]) by jaguar.mail.utk.edu (8.13.3/8.13.3) with ESMTP id j18DkI6n018261 for < ccl.net>; Tue, 8 Feb 2005 08:46:22 -0500 Message-ID: <4208C2A9.4000903 >< utk.edu> Date: Tue, 08 Feb 2005 08:46:17 -0500 From: Robert Hinde < utk.edu> Reply-To: rhinde >< utk.edu Organization: Univ. of Tennessee Chemistry Dept. User-Agent: Mozilla Thunderbird 0.7.3 (Macintosh/20040803) X-Accept-Language: en-us, en MIME-Version: 1.0 To: CHEMISTRY >< ccl.net Subject: CCL: Philosophy of Chemistry meeting Aug. 2005 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Status: No, hits=0.0 required=7.5 tests=none autolearn=no version=2.61 X-Scanned-By: MIMEDefang 2.45 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This meeting may be of interest to CCL members, particularly those currently discussing the reality of orbitals. Please direct questions to the conference chair, Jeffrey Kovac, not to me! Robert Hinde Univ. of Tennessee ----- Announcement and Call for Papers International Society for the Philosophy of Chemistry Ninth Summer Symposium August 7 - 11, 2005 University of Tennessee Knoxville, TN The ISPC Summer Symposium invites scholars to contribute papers in the general area of the philosophy of chemistry, including the history of chemistry and philosophical aspects of chemical education, as well as closely related subjects. Abstracts for papers to be presented at this symposium should be submitted to the symposium chair by May 1, 2005. Decisions concerning acceptance will be made as the papers are submitted. Information concerning abstract submission, registration, housing, and other symposium information can be found at www.chem.utk.edu/~ISPC. This website is currently being developed but will be available by March 1. Questions can be directed to the symposium chair, Jeffrey Kovac, by e-mail jkovac >< utk.edu. From chemistry-request@ccl.net Tue Feb 8 03:59:33 2005 Received: from mail.interware.hu (mail.interware.hu [195.70.32.130]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j188xU5J028149 for ; Tue, 8 Feb 2005 03:59:31 -0500 Received: from caracas-1511.adsl.interware.hu ([213.178.105.231] helo=cthulhu) by mail.interware.hu with smtp (Exim 4.43 #1 (Debian)) id 1CyQBm-00051q-Es; Tue, 08 Feb 2005 08:53:38 +0100 Message-ID: <002601c50db3$dadd7220$6401a8c0@cthulhu> From: "Tamas E. Gunda" To: "Steve Bowlus" , References: Subject: Re: CCL:CPU usage Date: Tue, 8 Feb 2005 08:57:26 +0100 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0023_01C50DBC.362B5770" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1478 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Spam-Status: No, hits=4.4 required=7.5 tests=HTML_30_40,HTML_FONTCOLOR_BLUE, HTML_MESSAGE,MK_YAHOO_REDIR_01,MY_DSL,MY_HTML_OBFU,WLS_URI_OPT_150 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------=_NextPart_000_0023_01C50DBC.362B5770 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I think it is not depending on the hardware, but the code: how it was = written, how it was compiled, how the threads are organized and used = etc. You can add more or less priority to a program/process from the = task manager, but it influences only the time sharing of the running = processes: a low priority process can run quick too (such as the Seti), = when no other big applications are used simultaneously, and vice versa. = On the other hand, the CPU usage heavily depends on what the program is = actually doing. For example when some geometry optimization is done on a = PC by Gamess using ab initio method, the displayed CPU usage values are = between 10 and 99% .----- Original Message -----=20 From: Steve Bowlus=20 To: Rodrigo Pontes ; chemistry {} ccl.net=20 Sent: Tuesday, February 08, 2005 4:08 AM Subject: CCL:CPU usage I have observed the same behavior with another quantum code on a = machine with the same CPU/OS combination. There was an earlier post to = this list asking about what seemed to be a similar problem with the Xeon = HT.=20 I tentatively interpret the behavior that the job is given 100% of one = of the HT execution streams, while the other stream is largely kept = idle. I have not found a way to tune this behavior. If you look at = ControlPanels >AdministrativeTools >ComputerManagement >Processors, you = will see there are two CPUs shown. These can be uninstalled, but I = haven't tried uninstalling one. Presumably, all CPU resources would be = concentrated in a single execution stream, but I haven't found any = documentation on this point. I also haven't found anything to suggest = the second "CPU" can be _re_installed. =20 I haven't talked to Intel or HP (who built my machine), either of whom = could shed light on all these issues. Seems a pity that information on = such a feature (or limitation) of HT technology would not be made = readily available. sb -----Original Message----- From: Computational Chemistry List = [mailto:chemistry-request {} ccl.net]On Behalf Of Rodrigo Pontes Sent: Monday, February 07, 2005 9:29 AM To: chemistry {} ccl.net Subject: CCL:CPU usage Dear Colleagues, We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz = machine. We observed (in the Task manager / performance) that the CPU = usage is only around 60%. Have someone noticed the same problem? How = could us set the configuration to a 100% CPU usage? Thanks in advance, Rodrigo M. Pontes Departemanto de Quimica Universidade Estadual de Maringa Brazil -------------------------------------------------------------------------= --- Yahoo! Acesso Gratis - Internet rapida e gratis. Instale o discador = do Yahoo! agora. ------=_NextPart_000_0023_01C50DBC.362B5770 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

I think it is not depending on the = hardware, but=20 the code: how it was written, how it was compiled, how the threads are = organized=20 and used etc. You can add more or less priority to a program/process = > from the=20 task manager, but it influences only the time sharing of the running = processes:=20 a low priority process can run quick too (such as the = Seti), when=20 no other big applications are used simultaneously, and vice=20 versa. On the other hand, the CPU usage heavily depends on what the = program=20 is actually doing. For example when some geometry optimization is = done on a=20 PC by Gamess using ab initio method, the displayed CPU usage values = are=20 between 10 and 99%

.----- Original Message -----
From:=20 Steve=20 Bowlus
To: Rodrigo Pontes ; chemistry {} ccl.net=20

Sent: Tuesday, February 08, = 2005 4:08=20 AM

Subject: CCL:CPU usage

I=20 have observed the same behavior with another quantum code on a machine = with=20 the same CPU/OS combination. There was an earlier post to this = list=20 asking about what seemed to be a similar problem with the Xeon HT.=20

I=20 tentatively interpret the behavior that the job is given 100% of one = of the HT=20 execution streams, while the other stream is largely kept idle. = I have=20 not found a way to tune this behavior. If you look at ControlPanels=20 >AdministrativeTools >ComputerManagement >Processors, = you will=20 see there are two CPUs shown. These can be uninstalled, but I = haven't=20 tried uninstalling one. Presumably, all CPU resources would be=20 concentrated in a single execution stream, but I haven't found any=20 documentation on this point. I also haven't found anything to = suggest=20 the second "CPU" can be _re_installed.

I=20 haven't talked to Intel or HP (who built my machine), either of whom = could=20 shed light on all these issues. Seems a pity that information on = such a=20 feature (or limitation) of HT technology would not be made readily=20 available.

sb

-----Original = Message-----
From:=20 Computational Chemistry List [mailto:chemistry-request {} ccl.net]On = Behalf Of=20 Rodrigo Pontes
Sent: Monday, February 07, 2005 9:29=20 AM
To: chemistry {} ccl.net
Subject: CCL:CPU=20 usage

Dear Colleagues,

We are running Gaussian on a Windows XP/Pentiun-IV HT=20 3.2GHz machine. We observed (in the Task manager / performance) = that=20 the CPU usage is only around 60%. Have someone noticed the = same=20 problem? How could us set the configuration to a 100% CPU = usage?

Thanks in advance,

Rodrigo M. Pontes

Departemanto de Quimica

Universidade Estadual de Maringa

Brazil

Yahoo!=20 Acesso Gratis - Internet rapida e gratis. Instale o discador do = Yahoo!=20 agora.

------=_NextPart_000_0023_01C50DBC.362B5770-- From chemistry-request@ccl.net Tue Feb 8 08:31:43 2005 Received: from web30807.mail.mud.yahoo.com (web30807.mail.mud.yahoo.com [68.142.200.150]) by server.ccl.net (8.12.8/8.12.8) with SMTP id j18DVei8014225 for ; Tue, 8 Feb 2005 08:31:40 -0500 Received: (qmail 73742 invoked by uid 60001); 8 Feb 2005 13:31:39 -0000 Comment: DomainKeys? See http://antispam.yahoo.com/domainkeys DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; b=HQp3T53eGrVcnE6zku9SXZ+Q7bE8NwJZg/EtCcAxitmc58pHhiIF+cCvrzG4YBtsbmdW310TItKhx2B4H7okOrFcDwzYJCFrlKdJKslwFfVklE9XBC+JYeUgsIOjlDozV1s9PlrpAql1+o++3nlco4WSWYI18OnNwiOQY1suBp4= ; Message-ID: <20050208133139.73740.qmail [a] web30807.mail.mud.yahoo.com> Received: from [68.46.0.3] by web30807.mail.mud.yahoo.com via HTTP; Tue, 08 Feb 2005 05:31:39 PST Date: Tue, 8 Feb 2005 05:31:39 -0800 (PST) From: Guosheng Wu Subject: Re: CCL:orbitals and reality To: Sengen Sun , chemistry [a] ccl.net In-Reply-To: <20050207072901.79222.qmail [a] web51301.mail.yahoo.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Spam-Status: No, hits=7.2 required=7.5 tests=FORGED_YAHOO_RCVD, FROM_ENDS_IN_NUMS,FROM_HAS_ULINE_NUMS,FVGT_m_MULTI_ODD2, RCVD_IN_DYNABLOCK,RCVD_IN_SORBS autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net --- Sengen Sun wrote: > The evident problem here to me is that some people > support the Nature paper based on how useful the > single WFs are in the mathematical treatments. This is > especially evident in the posting by Guosheng Wu. I do > not wish to address it separately. It was you who has been trying to make a number of statements on ANY kind of wave functions in many of your posts. Just pick a few recent cases: >In contrast, WFs disappear like ghosts outside our >mathemtical schemes! In fact, the 1954 Nobel prize was >awarded partially to Born for his interpretation that >WFs have no physical meaning. ( All WFS are like ghosts; and you changed the Nobel Prize announcement.) >I must again emphasize that the electronic structure >of a quantum system cannot function physically >according to WFs in chemical reactions. To my best >knowledge, I am the first person who explicitly states >so in the public, although this point was implied in >several occassions by Dewar, Bader, et. al. I also >wish to emphasize that chemistry is not about orbitals >but electron density and its reorganization as nuclei >change their positions. The third important point is >that chemical reactivity is decided by the easiness of >electronic reorganization. These three points are very >valuable in teaching and in scientific research, as >recognized by two CCL friends recently. >I have to say honestly that the authors of the Nature >papers and some people blindly supporting them are >scientifically naive on this particular issue and need >to re-study or re-think of QM. There are no physical >entities or physical properties directly associated >with WFs! .... ..... >All the mathematical schemes for the WF or orbital observation >are derived from our mental forces instead of any >meaningful physical interaction. Consequently, a >picture in human mind is obtained from forceless >mathematical schemes. This is not a scientific measurement! ... ... You have been trying to generalize some kind of general principles on any kind of wave functions, so the most effective approach to answer that is to pull out some very simple systems, accurately solvable, well defined, no confusion at all. If you really think that you have many discoveries that deserve respect > from REAL theoretical chemists in the world, please not only discuss your ideas at CCL, since this is just CCL, not TCL(theoretical chemistry list), not TPL(theoretical physics list) either, BUT publish it into some peer-reviewed scientific journals, such as Science, JACS or JCP or whatever, if someone would think that Nature always publishes non-scientific results. Does it make any sense when people are seeing exploding number of electronic publications without any kind of review? Sincerely, -Guosheng __________________________________ Do you Yahoo!? Meet the all-new My Yahoo! - Try it today! http://my.yahoo.com From chemistry-request@ccl.net Tue Feb 8 09:17:52 2005 Received: from mailserv.unb.ca (mailserv.unb.ca [131.202.3.23]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18EHoi8019728 for ; Tue, 8 Feb 2005 09:17:50 -0500 Received: from smeagol.unb.ca (smeagol.unb.ca [131.202.3.52]) by mailserv.unb.ca (8.13.1/8.13.1) with ESMTP id j18Corn8013737; Tue, 8 Feb 2005 08:50:53 -0400 Received: from smeagol.unb.ca (localhost.localdomain [127.0.0.1]) by smeagol.unb.ca (8.12.11/8.12.8) with ESMTP id j18CosBC022826; Tue, 8 Feb 2005 08:50:54 -0400 Received: (from apache@localhost) by smeagol.unb.ca (8.12.11/8.12.8/Submit) id j18CosBi022824; Tue, 8 Feb 2005 08:50:54 -0400 Received: from 131.202.181.4 ([131.202.181.4]) by webmail.unb.ca (IMP) with HTTP for ; Tue, 8 Feb 2005 08:50:54 -0400 Message-ID: <1107867054.4208b5ae35b7b(at)webmail.unb.ca> Date: Tue, 8 Feb 2005 08:50:54 -0400 From: Joe Harriman Reply-to: j.harriman(at)unb.ca To: tminagaw(at)purdue.edu Cc: chemistry(at)ccl.net Subject: Re: CCL:Making A Protonated Ligand with ADT References: <1107800390.4207b14651fa0(at)webmail.purdue.edu> In-Reply-To: <1107800390.4207b14651fa0(at)webmail.purdue.edu> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.2 X-Originating-IP: 131.202.181.4 X-UNB-MailScanner-Information: Please contact the ISP for more information X-UNB-VirusScanner: Found to be clean X-UNB-SpamDetails: X-MailScanner-From: s808o(at)unb.ca X-Spam-Status: No, hits=3.0 required=7.5 tests=NO_RDNS2 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Hi Taisuke, When you import your ligand into ADT the charges should not be affected. In the grid section you can choose to keep your original charges or use Gasteiger charges. If you want to use your charges (precalculated) you have to use a file format that saves partial charge data. A pdb file only contains information about atom types and bonding. As such ADT will automatically give your structure partial charges and if you do not have an ionization of +1 on the atom in question it will by default set the overall net charge to 0. The method that our group uses is to generate the partial charges (either by quantum or molecular mechanics) and save the ligand as a mol2 file. This is compatible with ADT and will allow you to import your own charges. I would try to generate the partial charges (with the proper protonation on the ligand), then save the molecule as a mol2 file and import into ADT. This should solve your problem. -Joe ---------------------------------------- D. Joseph Harriman PhD Candidate Dept. of Chemistry University of New Brunswick Room 228, Toole Hall ---------------------------------------- Quoting tminagaw(at)purdue.edu: > > "How to make A Protonated Ligand" > > I am using ADT(Autodock Tools) with Autodock 3.0.5 > I would like to know the way to make a "protoneted ligand" > pdbq file. > > As you can see, benA.pdbq in Autodock 3.0.5 example(3PTB) > has a total charge of 1.001, and I was trying to > reproduce the same file with an original 3TB.pdb downloaded > from Protein Data Bank. > > I extracted the ligand part (benA) from 3PTB.pdb, > added Hydrogens and read this as a ligand molecule using ADT. > However, ADT has assigned partial charges (Gasteiger charge) with > a net "total" charge of zero automatically. I need to assign a > positive (+1) total charge to the ligand. How do I request > ADT to assign non-zero ligand total charges (e.g., +1 or -1)? > Thank you. > > Taisuke Minagawa > Department of Physics > Purdue University, USA > > > > > -= This is automatically added to each message by the mailing script =- > To send e-mail to subscribers of CCL put the string CCL: on your Subject: line > and send your message to: CHEMISTRY(at)ccl.net > > Send your subscription/unsubscription requests to: CHEMISTRY-REQUEST(at)ccl.net > HOME Page: http://www.ccl.net | Jobs Page: http://www.ccl.net/jobs > > If your mail is bouncing from CCL.NET domain send it to the maintainer: > Jan Labanowski, jlabanow(at)nd.edu (read about it on CCL Home Page) > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ > > > > > > From chemistry-request@ccl.net Tue Feb 8 08:29:21 2005 Received: from ns.bme.hu (ns.bme.hu [152.66.116.1]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18DTJi8013684 for ; Tue, 8 Feb 2005 08:29:20 -0500 Received: from localhost (localhost [127.0.0.1]) by ns.bme.hu (Postfix) with ESMTP id 4B3A7106E96 for ; Tue, 8 Feb 2005 14:29:18 +0100 (CET) Received: from ns.bme.hu ([127.0.0.1]) by localhost (ns.bme.hu [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 16233-01-2 for ; Tue, 8 Feb 2005 14:29:18 +0100 (CET) Received: from wide (the.inc.bme.hu [152.66.63.76]) by ns.bme.hu (Postfix) with ESMTP id 06085106E83 for ; Tue, 8 Feb 2005 14:29:17 +0100 (CET) From: "Gabor I. Csonka" To: Subject: CCL: orbitals and reality Date: Tue, 8 Feb 2005 14:29:24 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-2" X-Mailer: Microsoft Office Outlook, Build 11.0.6353 In-Reply-To: <3.0.1.32.20050207140152.00f34cb0*at*iut-mel.ujf-grenoble.fr> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Thread-Index: AcUNMBwKdD7hocMLT/es/MQgYPSdHgAr0uXg Message-Id: <20050208132917.06085106E83*at*ns.bme.hu> X-Spam-Status: No, hits=3.0 required=7.5 tests=NO_RDNS2 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id j18DTLi8013703 I am not sure that everybody who is interested in this orbitals and reality discussion read the recent paper DOI: 10.1103/PhysRevLett.93.213001: Realization of Quantum Chemistry without Wave Functions through First-Order Semidefinite Programming David A. Mazziotti I cite the first few sentences of the introduction: "At a conference in 1959 Coulson proposed ''banishing'' the wave function from electronic structure calculations in quantum chemistry [1]. Lvwdin and Mayer [2] had shown earlier that the electronic energies and properties of atoms and molecules are computable simply > from an effective two-electron density distribution known as the two-electron reduced density matrix (2-RDM)." [1] A. J. Coleman and V. I. Yukalov, Reduced Density Matrices: Coulson's Challenge (Springer-Verlag, New York, 2000). [2] P.O. Lvwdin, Phys. Rev. 97, 1474 (1955); J. E. Mayer, Phys. Rev. 100, 1579 (1955). I think it is an interesting contrast to the Nature paper. For me the Phys. Rev. Lett. Paper is quite reasonable. Interestingly N2 is taken as one of the two examples. Best regards, Gabor I. Csonka From chemistry-request@ccl.net Tue Feb 8 10:10:44 2005 Received: from web53507.mail.yahoo.com (web53507.mail.yahoo.com [206.190.37.68]) by server.ccl.net (8.12.8/8.12.8) with SMTP id j18FAgi8025560 for ; Tue, 8 Feb 2005 10:10:42 -0500 Received: (qmail 49898 invoked by uid 60001); 8 Feb 2005 14:10:41 -0000 Comment: DomainKeys? See http://antispam.yahoo.com/domainkeys DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; b=WdabwXjDTpBoyUPBhe+e+bozOtfSlOwBdrzoglA45+uRxn0kPFcPvcZWwEkktylhRlmjghqbY+c+NKrDW9oTgf3SG/92vsc6vZ3LtNhOZ2lRVlhfmctfGwri05cUfQ1RGfjcyw5HBZX4m+SM5oqceQ+aMO/FCTspWK2SAXfFg0E= ; Message-ID: <20050208141041.49896.qmail-.at.-web53507.mail.yahoo.com> Received: from [148.234.75.162] by web53507.mail.yahoo.com via HTTP; Tue, 08 Feb 2005 06:10:41 PST Date: Tue, 8 Feb 2005 06:10:41 -0800 (PST) From: Victor Manuel Rosas-Garcia Subject: Re: CCL:CPU usage To: chemistry-.at.-ccl.net Cc: Steve Bowlus In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Spam-Status: No, hits=3.0 required=7.5 tests=FROM_ENDS_IN_NUMS, MAILTO_TO_SPAM_ADDR autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net I had a similar problem when running GAMESS under Fedora Core 2. Then I disabled hyperthreading, from the setup menu upon starting the machine, and cpu usage went up to 100%. When I installed the OS, the Fedora Core 2 installer setup two kernels, one for 1 CPU and the SMP variety. After disabling hyperthreading I had to make sure the machine always runs the kernel for 1 CPU so that I get 100% CPU usage. --- Steve Bowlus wrote: > I have observed the same behavior with another quantum code on a machine > with the same CPU/OS combination. There was an earlier post to this list > asking about what seemed to be a similar problem with the Xeon HT. > > I tentatively interpret the behavior that the job is given 100% of one of > the HT execution streams, while the other stream is largely kept idle. I > have not found a way to tune this behavior. If you look at ControlPanels > >AdministrativeTools >ComputerManagement >Processors, you will see there are > two CPUs shown. These can be uninstalled, but I haven't tried uninstalling > one. Presumably, all CPU resources would be concentrated in a single > execution stream, but I haven't found any documentation on this point. I > also haven't found anything to suggest the second "CPU" can be > _re_installed. > > I haven't talked to Intel or HP (who built my machine), either of whom could > shed light on all these issues. Seems a pity that information on such a > feature (or limitation) of HT technology would not be made readily > available. > > sb > > > > -----Original Message----- > From: Computational Chemistry List [mailto:chemistry-request-.at.-ccl.net]On > Behalf Of Rodrigo Pontes > Sent: Monday, February 07, 2005 9:29 AM > To: chemistry-.at.-ccl.net > Subject: CCL:CPU usage > > > Dear Colleagues, > > We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz machine. We > observed (in the Task manager / performance) that the CPU usage is only > around 60%. Have someone noticed the same problem? How could us set the > configuration to a 100% CPU usage? > > Thanks in advance, > > Rodrigo M. Pontes > Departemanto de Quimica > Universidade Estadual de Maringa > Brazil ===== Victor M. Rosas Garcma Theoretical Chemistry Apprentice e-mail: quimico69-.at.-yahoo.com __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From chemistry-request@ccl.net Tue Feb 8 05:00:34 2005 Received: from lgdx04.lg.ehu.es (lgdx04.lg.ehu.es [158.227.1.36]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18A0Ci8009630 for ; Tue, 8 Feb 2005 05:00:21 -0500 Received: by lgdx04.lg.ehu.es id JAA0000022433; Tue, 8 Feb 2005 09:46:04 +0100 (MET) Message-ID: <007801c50dba$a1611720$4a2ee39e@lcpycx> From: "Pablo Vitoria" To: "Steve Bowlus" , "Rodrigo Pontes" , References: Subject: Re: CCL:CPU usage Date: Tue, 8 Feb 2005 09:45:32 +0100 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_006E_01C50DC2.EE92C5E0" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2527 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2527 X-Spam-Status: No, hits=3.6 required=7.5 tests=HTML_30_40,HTML_FONTCOLOR_BLUE, HTML_MESSAGE,MK_YAHOO_REDIR_01,MY_HTML_OBFU,WLS_URI_OPT_150 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------=_NextPart_000_006E_01C50DC2.EE92C5E0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, I don=B4t have much experience with G03 under Windows, but if it is = really a HT problem, you could try disabling HyperThreading in the BIOS = and see what happens. I run G03 under Linux in a dual Xeon, and I = disabled HT. Pablo ----- Original Message -----=20 From: Steve Bowlus=20 To: Rodrigo Pontes ; chemistry[at]ccl.net=20 Sent: Tuesday, February 08, 2005 4:08 AM Subject: CCL:CPU usage I have observed the same behavior with another quantum code on a = machine with the same CPU/OS combination. There was an earlier post to = this list asking about what seemed to be a similar problem with the Xeon = HT.=20 I tentatively interpret the behavior that the job is given 100% of one = of the HT execution streams, while the other stream is largely kept = idle. I have not found a way to tune this behavior. If you look at = ControlPanels >AdministrativeTools >ComputerManagement >Processors, you = will see there are two CPUs shown. These can be uninstalled, but I = haven't tried uninstalling one. Presumably, all CPU resources would be = concentrated in a single execution stream, but I haven't found any = documentation on this point. I also haven't found anything to suggest = the second "CPU" can be _re_installed. =20 I haven't talked to Intel or HP (who built my machine), either of whom = could shed light on all these issues. Seems a pity that information on = such a feature (or limitation) of HT technology would not be made = readily available. sb -----Original Message----- From: Computational Chemistry List = [mailto:chemistry-request[at]ccl.net]On Behalf Of Rodrigo Pontes Sent: Monday, February 07, 2005 9:29 AM To: chemistry[at]ccl.net Subject: CCL:CPU usage Dear Colleagues, We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz = machine. We observed (in the Task manager / performance) that the CPU = usage is only around 60%. Have someone noticed the same problem? How = could us set the configuration to a 100% CPU usage? Thanks in advance, Rodrigo M. Pontes Departemanto de Quimica Universidade Estadual de Maringa Brazil -------------------------------------------------------------------------= --- Yahoo! Acesso Gratis - Internet rapida e gratis. Instale o discador = do Yahoo! agora. ------=_NextPart_000_006E_01C50DC2.EE92C5E0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Hi,

I don=B4t have much experience with G03 = under=20 Windows, but if it is really a HT problem, you could try disabling=20 HyperThreading in the BIOS and see what happens. I run G03 under Linux = in a dual=20 Xeon, and I disabled HT.

Pablo

----- Original Message -----
From:=20 Steve=20 Bowlus
To: Rodrigo Pontes ; chemistry[at]ccl.net=20

Sent: Tuesday, February 08, = 2005 4:08=20 AM

Subject: CCL:CPU usage

I=20 have observed the same behavior with another quantum code on a machine = with=20 the same CPU/OS combination. There was an earlier post to this = list=20 asking about what seemed to be a similar problem with the Xeon HT.=20

I=20 tentatively interpret the behavior that the job is given 100% of one = of the HT=20 execution streams, while the other stream is largely kept idle. = I have=20 not found a way to tune this behavior. If you look at ControlPanels=20 >AdministrativeTools >ComputerManagement >Processors, = you will=20 see there are two CPUs shown. These can be uninstalled, but I = haven't=20 tried uninstalling one. Presumably, all CPU resources would be=20 concentrated in a single execution stream, but I haven't found any=20 documentation on this point. I also haven't found anything to = suggest=20 the second "CPU" can be _re_installed.

I=20 haven't talked to Intel or HP (who built my machine), either of whom = could=20 shed light on all these issues. Seems a pity that information on = such a=20 feature (or limitation) of HT technology would not be made readily=20 available.

sb

-----Original = Message-----
From:=20 Computational Chemistry List [mailto:chemistry-request[at]ccl.net]On = Behalf Of=20 Rodrigo Pontes
Sent: Monday, February 07, 2005 9:29=20 AM
To: chemistry[at]ccl.net
Subject: CCL:CPU=20 usage

Dear Colleagues,

We are running Gaussian on a Windows XP/Pentiun-IV HT=20 3.2GHz machine. We observed (in the Task manager / performance) = that=20 the CPU usage is only around 60%. Have someone noticed the = same=20 problem? How could us set the configuration to a 100% CPU = usage?

Thanks in advance,

Rodrigo M. Pontes

Departemanto de Quimica

Universidade Estadual de Maringa

Brazil

Yahoo!=20 Acesso Gratis - Internet rapida e gratis. Instale o discador do = Yahoo!=20 agora.

------=_NextPart_000_006E_01C50DC2.EE92C5E0-- From chemistry-request@ccl.net Tue Feb 8 12:00:39 2005 Received: from neptune.tranzyme.com (neptune.tranzyme.com [132.210.158.201]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18H0ci8015221 for ; Tue, 8 Feb 2005 12:00:38 -0500 Received: from mercure.tranzyme.com (dcsher01.tranzyme.com [10.32.2.3]) by neptune.tranzyme.com (8.12.10/8.12.10) with ESMTP id j18FHBic017224 for ; Tue, 8 Feb 2005 10:17:11 -0500 Content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C50DF1.5E8D64B2" Subject: CCL: EF optimization in MOPAC 7 X-MIMEOLE: Produced By Microsoft Exchange V6.5.7226.0 Date: Tue, 8 Feb 2005 10:17:56 -0500 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: CCL: EF optimization in MOPAC 7 thread-index: AcUN8V4jXrbvChLYQm2r70WHx+i/6A== From: "Axel Mathieu" To: X-Scanned-By: MIMEDefang 2.48 on 10.32.2.1 X-Spam-Status: No, hits=0.0 required=7.5 tests=HTML_MESSAGE autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------_=_NextPart_001_01C50DF1.5E8D64B2 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear CCL'ers, =20 I've been running some AM1 calculations and whenever EF is used it never = performs the optimization for more than 100 cycles. I've tried = increasing ITRY and CYCLE but it still stays stuck at the default 100 = cycles. Am I missing something here? Is this a bug in MOPAC 7? =20 Here are the keywords used: XYZ MMOK AM1 T=3D1D LET GNORM=3D0.01 NOINTER = DUMP=3D60 EF ITRY=3D1000 Thanks for your help ... =20 APM =20 ------_=_NextPart_001_01C50DF1.5E8D64B2 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Dear CCL’ers,

I’ve been running some AM1 calculations and = whenever EF is used it never performs the optimization for more than 100 cycles. = I’ve tried increasing ITRY and CYCLE but it still stays stuck at the default = 100 cycles. Am I missing something here? Is this a bug in MOPAC = 7?

Here are the keywords used: XYZ MMOK AM1 T=3D1D = LET GNORM=3D0.01 NOINTER DUMP=3D60 EF = ITRY=3D1000

Thanks for your help = …

APM

------_=_NextPart_001_01C50DF1.5E8D64B2-- From chemistry-request@ccl.net Tue Feb 8 14:50:03 2005 Received: from doppler.zen.co.uk (doppler.zen.co.uk [212.23.3.27]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18Jo1i8014643 for ; Tue, 8 Feb 2005 14:50:01 -0500 Received: from [212.23.3.141] (helo=heisenberg.zen.co.uk) by doppler.zen.co.uk with esmtp (Exim 4.34) id 1CyZkA-0008Cb-Rp for chemistry :a: ccl.net; Tue, 08 Feb 2005 18:05:46 +0000 Received: from [82.68.46.238] (helo=cresset23) by heisenberg.zen.co.uk with esmtp (Exim 4.30) id 1CyZk9-0008Ub-90 for chemistry :a: ccl.net; Tue, 08 Feb 2005 18:05:45 +0000 From: "Mark Mackey" To: Subject: RE: CPU usage Date: Tue, 8 Feb 2005 18:07:46 -0000 Message-ID: <00ad01c50e09$17f73a60$0207a8c0@cresset23> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook, Build 10.0.6626 Importance: Normal In-Reply-To: <20050208141041.49896.qmail :a: web53507.mail.yahoo.com> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Originating-Heisenberg-IP: [82.68.46.238] X-Spam-Status: No, hits=1.5 required=7.5 tests=MAILTO_TO_SPAM_ADDR autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id j18Jo3i8014651 I think there's some confusion here as to the effect of hyperthreading. When hyperthreading is enabled, the CPU appears to the operating system as though it was two CPUs, and some clever on-chip trickery allows one of these 'virtual cpus' to keep processing when the other one stalls (due to eg a memory cache miss). If you have only one process running, it runs in one of the 'virtual CPUs' at exactly the same speed that it would run if HT was turned off. The benefit of HT is that if there is more than one process running, you get slightly faster throughput than if HT was not enabled. Note that this benefit is generally only slight (5-10% max). Effectively, all that HT gives you is the ability of the CPU to 'context switch' to a second task very quickly if the current task stalls: there is no real extra CPU power made available by HT for a single execution stream. One thing to remember is that although the OS thinks there are two CPUs there, there is really only one. The 'CPU utilisation' monitors in windows (and 'top' in Linux etc) get a little confused by this. In the original poster's case, you have a Windows machine running Gaussian at ~100% CPU utilisation, and assorted Windows background processes also running using a few cycles here and there. With HT turned off, the machine thinks it has only one processor, that processor is doing stuff all of the time, and so it reports 100% CPU utilisation. With HT turned on, Gaussian runs at 100% of one 'virtual CPU', the Windows background processors all run on the other, and Windows thinks that there are two CPUs, one fully loaded, the other not doing much, so reports 55% CPU utilisation. In both cases Gaussian is running as fast as it can and will finish the same jobs in more-or-less the same time: my guess is that the 'HT on' case will go 1% faster or so (as Gaussian is competing less with the Windows background stuff). The symptoms of "Why is it only using 60% of my processing power" are purely due to the Windows tools not fully understanding the underlying hardware: they do NOT indicate that the CPU is in reality being under-utilised. It's a bug/misunderstanding in the CPU monitor tool, not a problem with the calculation or the system as a whole. The only way to get 100% displayed CPU utilisation is to run two Gaussians simultaneously, in which case both will take very slightly under twice as long to finish[1] but the CPU monitor will be happy that 'both' CPUs are being used fully. [1] Well, probably more than twice as long in the real world because of contention for memory and disk resources and so on. -- Mark Mackey Cresset Biomolecular Discovery -----Original Message----- From: Computational Chemistry List [mailto:chemistry-request :a: ccl.net] On Behalf Of Victor Manuel Rosas-Garcia Sent: 08 February 2005 14:11 To: chemistry :a: ccl.net Cc: Steve Bowlus Subject: CCL:CPU usage I had a similar problem when running GAMESS under Fedora Core 2. Then I disabled hyperthreading, from the setup menu upon starting the machine, and cpu usage went up to 100%. When I installed the OS, the Fedora Core 2 installer setup two kernels, one for 1 CPU and the SMP variety. After disabling hyperthreading I had to make sure the machine always runs the kernel for 1 CPU so that I get 100% CPU usage. --- Steve Bowlus wrote: > I have observed the same behavior with another quantum code on a machine > with the same CPU/OS combination. There was an earlier post to this list > asking about what seemed to be a similar problem with the Xeon HT. > > I tentatively interpret the behavior that the job is given 100% of one of > the HT execution streams, while the other stream is largely kept idle. I > have not found a way to tune this behavior. If you look at ControlPanels > >AdministrativeTools >ComputerManagement >Processors, you will see there are > two CPUs shown. These can be uninstalled, but I haven't tried uninstalling > one. Presumably, all CPU resources would be concentrated in a single > execution stream, but I haven't found any documentation on this point. I > also haven't found anything to suggest the second "CPU" can be > _re_installed. > > I haven't talked to Intel or HP (who built my machine), either of whom could > shed light on all these issues. Seems a pity that information on such a > feature (or limitation) of HT technology would not be made readily > available. > > sb > > > > -----Original Message----- > From: Computational Chemistry List [mailto:chemistry-request :a: ccl.net]On > Behalf Of Rodrigo Pontes > Sent: Monday, February 07, 2005 9:29 AM > To: chemistry :a: ccl.net > Subject: CCL:CPU usage > > > Dear Colleagues, > > We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz machine. We > observed (in the Task manager / performance) that the CPU usage is only > around 60%. Have someone noticed the same problem? How could us set the > configuration to a 100% CPU usage? > > Thanks in advance, > > Rodrigo M. Pontes > Departemanto de Quimica > Universidade Estadual de Maringa > Brazil ===== Victor M. Rosas Garcma Theoretical Chemistry Apprentice e-mail: quimico69 :a: yahoo.com __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -= This is automatically added to each message by the mailing script =- To send e-mail to subscribers of CCL put the string CCL: on your Subject: line and send your message to: CHEMISTRY :a: ccl.net Send your subscription/unsubscription requests to: CHEMISTRY-REQUEST :a: ccl.net HOME Page: http://www.ccl.net | Jobs Page: http://www.ccl.net/jobs If your mail is bouncing from CCL.NET domain send it to the maintainer: Jan Labanowski, jlabanow :a: nd.edu (read about it on CCL Home Page) -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ From chemistry-request@ccl.net Tue Feb 8 13:26:58 2005 Received: from imgate1.cshore.com (imgate1.cshore.com [209.113.151.8]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18IQti8031370 for ; Tue, 8 Feb 2005 13:26:55 -0500 Received: from mailsrv.lorentzian.com (unknown [65.113.125.132]) by imgate1.cshore.com (Postfix) with ESMTP id A0D45AB49 for ; Tue, 8 Feb 2005 12:27:30 -0500 (EST) Received: from svega.gaussian.com (unknown [193.0.9.49]) by mailsrv.lorentzian.com (Postfix) with ESMTP id 152E8C943 for ; Tue, 8 Feb 2005 12:18:21 -0500 (EST) Received: by svega.gaussian.com (Postfix, from userid 350) id C09FB272EE7; Tue, 8 Feb 2005 12:26:55 -0500 (EST) Date: Tue, 8 Feb 2005 12:26:55 -0500 To: chemistry/at/ccl.net Subject: CCL: Gaussian workshop March 29 - April 1, Univ. of Alabama Message-ID: <20050208172655.GA25203/at/svega.gaussian.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.6i From: frisch/at/gaussian.com (Michael Frisch) X-Spam-Status: No, hits=1.0 required=7.5 tests=NO_RDNS autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Gaussian, Inc. and the University of Alabama will be holding a workshop: Introduction to Gaussian: Theory and Practice March 29 - April 1, 2005 at the Tuscaloosa campus of the University of Alabama The workshop will cover the full range of methods available in the Gaussian 03 package with emphasis on new methods and features which make Gaussian applicable to an ever widening spectrum of research applications. The workshop is structured to provide an introduction to electronic structure theory as well as a hands-on review for researchers active in the field. It is open to researchers at all levels of academic, government and industrial research. Details on the contents of the workshop and registration are on our website: http://www.gaussian.com From chemistry-request@ccl.net Tue Feb 8 14:47:30 2005 Received: from mail-h12-03.cc.ksu.edu (mail-h12-03.cc.ksu.edu [129.130.12.152]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18JlSi8014408 for ; Tue, 8 Feb 2005 14:47:29 -0500 Received: from [129.130.125.202] ([129.130.125.202]) by mail-h12-03.cc.ksu.edu (8.12.9/8.12.9) with ESMTP id j18IkRU9004090; Tue, 8 Feb 2005 12:46:31 -0600 (CST) Message-ID: <42090903.6030601 !! ksu.edu> Date: Tue, 08 Feb 2005 12:46:27 -0600 From: Gustavo Seabra User-Agent: Mozilla Thunderbird 0.9 (X11/20041127) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Pablo Vitoria , CHEMISTRY !! ccl.net Subject: Re: CCL:CPU usage References: <007801c50dba$a1611720$4a2ee39e@lcpycx> In-Reply-To: <007801c50dba$a1611720$4a2ee39e@lcpycx> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit X-Virus-Scanned: ClamAV 0.80/700/Thu Feb 3 17:33:15 2005 clamav-milter version 0.80j on virusfilter2.cc.ksu.edu X-Virus-Status: Clean X-Spam-Status: No, hits=5.2 required=7.5 tests=MK_YAHOO_REDIR_01,NO_RDNS2, WLS_URI_OPT_150 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Pablo Vitoria wrote: > Hi, > > I don4t have much experience with G03 under Windows, but if it is > really a HT problem, you could try disabling HyperThreading in the > BIOS and see what happens. I run G03 under Linux in a dual Xeon, and I > disabled HT. > > Pablo > > ----- Original Message ----- > *From:* Steve Bowlus > *To:* Rodrigo Pontes ; > chemistry !! ccl.net > *Sent:* Tuesday, February 08, 2005 4:08 AM > *Subject:* CCL:CPU usage > > I have observed the same behavior with another quantum code on a > machine with the same CPU/OS combination. There was an earlier > post to this list asking about what seemed to be a similar problem > with the Xeon HT. > > I tentatively interpret the behavior that the job is given 100% of > one of the HT execution streams, while the other stream is largely > kept idle. I have not found a way to tune this behavior. If you > look at ControlPanels >AdministrativeTools >ComputerManagement > >Processors, you will see there are two CPUs shown. These can be > uninstalled, but I haven't tried uninstalling one. Presumably, > all CPU resources would be concentrated in a single execution > stream, but I haven't found any documentation on this point. I > also haven't found anything to suggest the second "CPU" can be > _re_installed. > > I haven't talked to Intel or HP (who built my machine), either of > whom could shed light on all these issues. Seems a pity that > information on such a feature (or limitation) of HT technology > would not be made readily available. > > sb > > > > -----Original Message----- > *From:* Computational Chemistry List > [mailto:chemistry-request !! ccl.net]*On Behalf Of *Rodrigo Pontes > *Sent:* Monday, February 07, 2005 9:29 AM > *To:* chemistry !! ccl.net > *Subject:* CCL:CPU usage > > Dear Colleagues, > > We are running Gaussian on a Windows XP/Pentiun-IV HT > 3.2GHz machine. We observed (in the Task manager / > performance) that the CPU usage is only around 60%. > Have someone noticed the same problem? How could us set the > configuration to a 100% CPU usage? > > Thanks in advance, > > Rodrigo M. Pontes > Departemanto de Quimica > Universidade Estadual de Maringa > Brazil > > ------------------------------------------------------------------------ > Yahoo! Acesso Gratis > > - Internet rapida e gratis. Instale o discador do Yahoo! agora. > AFAIK, the HT actually makes a second "virtual" CPU. In this case, wouldn't using %NProc=2 in the Gaussian input file help? If it works, that would likely be true for the Fedora version as well. -- ---------------------------------- Gustavo Seabra - Graduate Student Chemistry Department Kansas State University ---------------------------------- If at first you don't succeed... ...skydiving is not for you. From chemistry-request@ccl.net Tue Feb 8 13:37:20 2005 Received: from mailgw4.technion.ac.il (mailgw4.technion.ac.il [132.68.238.37]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18IbHi8001731 for >ccl.net>; Tue, 8 Feb 2005 13:37:18 -0500 Received: from localhost (localhost.localdomain [127.0.0.1]) by mailgw4.technion.ac.il (Postfix) with ESMTP id 72ACAF7986 for >ccl.net>; Tue, 8 Feb 2005 18:34:11 +0200 (IST) (envelope-from chrirena<>techunix.technion.ac.il) Received: from mailgw4.technion.ac.il ([127.0.0.1]) by localhost (mailgw4.technion.ac.il [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 31161-01-3 for >ccl.net>; Tue, 8 Feb 2005 18:34:11 +0200 (IST) Received: from techunix.technion.ac.il (techunix.technion.ac.il [132.68.1.28]) by mailgw4.technion.ac.il (Postfix) with ESMTP id 44F14F7982 for >ccl.net>; Tue, 8 Feb 2005 18:34:11 +0200 (IST) (envelope-from chrirena<>techunix.technion.ac.il) Received: from localhost (localhost [127.0.0.1]) by techunix.technion.ac.il (Postfix) with ESMTP id 06444356F for >ccl.net>; Tue, 8 Feb 2005 18:37:57 +0200 (IST) (envelope-from chrirena<>techunix.technion.ac.il) Received: from techunix.technion.ac.il ([127.0.0.1]) by localhost (techunix.technion.ac.il [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 25843-01-74 for >ccl.net>; Tue, 8 Feb 2005 18:37:56 +0200 (IST) Received: from zaika (zaika.technion.ac.il [132.68.80.38]) by techunix.technion.ac.il (Postfix) with SMTP id BA2533539 for >ccl.net>; Tue, 8 Feb 2005 18:37:56 +0200 (IST) (envelope-from chrirena<>tx.technion.ac.il) Message-ID: <049901c50dfc$cb5c7140$26504484@zaika> From: "Irena Efremenko" >techunix.technion.ac.il> To: "ccl" >ccl.net> Subject: SCIPCM solvation model in Gaussian Date: Tue, 8 Feb 2005 18:39:44 +0200 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0496_01C50E0D.8ED95A60" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 X-Virus-Scanned: by amavisd-new at technion.ac.il X-Virus-Scanned: by amavisd-new at technion.ac.il X-Spam-Status: No, hits=5.0 required=7.5 tests=HTML_40_50, HTML_FONTCOLOR_UNSAFE,HTML_MESSAGE,NO_RDNS2 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------=_NextPart_000_0496_01C50E0D.8ED95A60 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: quoted-printable Dear CCLers, I am trying to calculate the solvation of metal clusters with G03 = employing the SCIPCM method. It works OK for Pd clusters but for Cu = clusters, even for one-atomic Cu+ ion, I get the following error = message: A solvent integration origin lies outside the cavity! Here is my input file: #P Becke3LYP/Gen Pseudo=3DRead SCF(MaxCycle=3D250) SCRF(SCIPCM)=20 Cu+ 1 1 Cu Cu 0 LANL2DZ **** Cu 0 LANL2 78.39 0.0000004 =20 I tried all the options suggested in Gaussian manual and varied the = isodensity value from 0.4 to 0.0000004 and always got the same error = message. Similar question was posted to CCl in 2001 and 2002: CCL.NET 2001.03.10-002 CCL.NET 2002.09.03-010 But I did not find a positive answer. What is the problem here and how to overcome it? Thanks in advance for your help, Irena Efremenko --------------- Dr. Irena Efremenko Wolfson Department of Chemical Engineering Technion - Israel Institute of Technology Haifa 32000, Israel Phone: 972-4-8293561 Fax: 972-4-8295672 ------=_NextPart_000_0496_01C50E0D.8ED95A60 Content-Type: text/html; charset="windows-1255" Content-Transfer-Encoding: quoted-printable

Dear CCLers,

I am trying to calculate the = solvation of=20 metal clusters with G03 employing the SCIPCM method. It works OK for Pd = clusters=20 but for Cu clusters, even for one-atomic Cu+ ion, I get the following = error=20 message:

A solvent integration origin lies outside the=20 cavity!

Here is my input file:

#P Becke3LYP/Gen Pseudo=3DRead=20 SCF(MaxCycle=3D250)
SCRF(SCIPCM)

Cu+

1 1
Cu

Cu = 0
LANL2DZ
****

Cu 0
LANL2

78.39=20 0.0000004

I tried all the options = suggested in=20 Gaussian manual and varied the isodensity value from 0.4 to = 0.0000004 and=20 always got the same error message.

Similar question was posted to = CCl in 2001=20 and 2002:

CCL.NET 2001.03.10-002

CCL.NET 2002.09.03-010
But = I did not=20 find a positive answer.

What is the problem here and how = to overcome=20 it?

Thanks in advance for your=20 help,
Irena = Efremenko

---------------
Dr. Irena = Efremenko
Wolfson=20 Department of Chemical Engineering
Technion - Israel Institute of=20 Technology
Haifa 32000, Israel
Phone: 972-4-8293561
Fax:=20 972-4-8295672

------=_NextPart_000_0496_01C50E0D.8ED95A60-- From chemistry-request@ccl.net Tue Feb 8 13:29:14 2005 Received: from exmails2.chem.ucla.edu (exmails2.chem.ucla.edu [169.232.134.3]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18ITCi8031890 for ; Tue, 8 Feb 2005 13:29:13 -0500 Received: from [169.232.227.161] (ts18-148.dialup.bol.ucla.edu [169.232.227.161]) (authenticated bits=0) by exmails2.chem.ucla.edu (8.12.8/8.12.8) with ESMTP id j18ISfuh009335; Tue, 8 Feb 2005 10:28:59 -0800 Mime-Version: 1.0 X-Sender: scerri +*+ chlorine.chem.ucla.edu (Unverified) Message-Id: Date: Tue, 8 Feb 2005 10:28:31 -0800 To: PHILCHEM +*+ LISTSERV.SC.EDU From: Eric Scerri Subject: Ninth Summer Symposium/ Philosophy of chemistry Cc: history of chemistry , HOPOS-L +*+ listserv.nd.edu, CHEMED-L +*+ MAILER.UWF.EDU, CHEM-EDUCATION +*+ JISCMAIL.AC.UK, ccl Content-Type: text/plain; charset="us-ascii" ; format="flowed" X-Spam-Status: No, hits=0.0 required=7.5 tests=none autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Announcement and Call for Papers International Society for the Philosophy of Chemistry Ninth Summer Symposium August 7 - 11, 2005 University of Tennessee Knoxville, TN The ISPC Summer Symposium invites scholars to contribute papers in the general area of the philosophy of chemistry, including the history of chemistry and philosophical aspects of chemical education, as well as closely related subjects. Abstracts for papers to be presented at this symposium should be submitted to the symposium chair by May 1, 2005. Decisions concerning acceptance will be made as the papers are submitted. Information concerning abstract submission, registration, housing, and other symposium information can be found at www.chem.utk.edu/~ISPC. This website is currently being developed but will be available by March 1. Questions can be directed to the symposium chair, Jeffrey Kovac, by e-mail jkovac +*+ utk.edu. -- Dr. Eric Scerri , UCLA, Department of Chemistry & Biochemistry, 607 Charles E. Young Drive East, Los Angeles, CA 90095-1569 USA E-mail : scerri +*+ chem.ucla.edu tel: 310 206 7443 fax: 310 206 2061 Web Page: http://www.chem.ucla.edu/dept/Faculty/scerri/index.html Editor of Foundations of Chemistry http://www.kluweronline.com/issn/1386-4238 Also see International Society for the Philosophy of Chemistry http://www.georgetown.edu/earleyj/ISPC.html From chemistry-request@ccl.net Tue Feb 8 12:39:03 2005 Received: from avas4.olemiss.edu (avas4.olemiss.edu [130.74.120.14]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18Hd2i8023072 for ; Tue, 8 Feb 2005 12:39:02 -0500 Received: from olemiss.edu (webmail.olemiss.edu [130.74.120.12] (may be forged)) by avas4.olemiss.edu (8.12.11/8.12.11) with ESMTP id j18Gakmd016412 for ; Tue, 8 Feb 2005 10:36:46 -0600 (envelope-from rjd-.at.-olemiss.edu) Sender: rjd-.at.-olemiss.edu From: "rjd" Reply-to: rjd-.at.-olemiss.edu To: CHEMISTRY-.at.-ccl.net Subject: Symposium to Honor Michael L. Klein X-Mailer: Quality Web Email v3.1e, http://netwinsite.com/refw.htm Date: Tue, 08 Feb 2005 10:36:46 -0600 Message-id: <4208ea9e.30f.339c.1082834477-.at.-olemiss.edu> MIME-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" X-PMX-Version: 4.7.0.111621, Antispam-Engine: 2.0.2.0, Antispam-Data: 2005.2.8.3 X-Spam-Status: No, hits=0.0 required=7.5 tests=none autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id j18Hd4i8023086 Press Release February 8, 2005 From: Dr. Douglas Tobias and Dr. Robert Doerksen Subject: Michael Klein 65th Birthday Symposium Venue: American Chemical Society, COMP - Division of Computers in Chemistry 229th ACS National Meeting, San Diego, CA, March 13-17, 2005 Purpose of symposium: This symposium honors Professor Michael L. Klein, Hepburn Professor of Physical Science, Department of Chemistry, University of Pennsylvania (Philadelphia, PA, USA) and Director of the Laboratory for Research on the Structure of Matter also at University of Pennsylvania. Klein is a Fellow of the American Academy of Arts and Sciences, Fellow of the Royal Society (UK), and winner of the Rahman Prize (1999) and CECAM Prize (2004). He has made outstanding contributions to molecular simulation throughout his career, often collaborated with industry, is a leading proponent of science policy and supercomputer development, and has supervised the education of many students and postdoctoral fellows, who are honoring him with this symposium on the occasion of his 65th birthday. The American Chemical Society conference is a principal forum for the exchange of information and discussion on the forefronts of molecular simulation. The program format consists of 40 invited oral presentations. Sponsorship: Sponsorship funding from the following organizations is helping to ensure the success of this symposium: Lawrence Livermore National Laboratory, Chemistry and Materials Science, Livermore, CA IBM Research Division, Thomas J. Watson Research Center, Yorktown Heights, NY Colgate-Palmolive Technology Division, Piscataway, NJ Laboratory for Applied Drug Design and Synthesis (LADDS), Department of Medicinal Chemistry, University of Mississippi, Oxford, MS Session Details: The symposium will feature leaders in the field of molecular simulation, including many former Klein group members. Sessions are entitled: advances in computer simulation methods; theory and modeling of polymers, complex fluids, and glasses; simulation of biological ion channels and enzyme reactions; structure and dynamics of clusters, liquids, and materials; and biomolecular interactions: experiments and modeling. The symposium will highlight new methodological developments in molecular dynamics and Monte Carlo technology, non-Hamiltonian dynamics, coarse-graining, metadynamics, mixed quantum/classical methods, statistical mechanics, polarizable force fields, electron dynamics, linear scaling algorithms. Also featured will be applications of molecular dynamics (MD) and Monte Carlo simulations, Car-Parrinello MD, ab initio MD, density-functional theory and QM/MM. The diversity of physical systems covered is remarkable, including structured liquids, glasses, ionic liquids, liquid interfaces, bilayers, colloids, micelles, clusters, nanoparticles, solids, surfaces, proteins, ion channels, enzymes, polymers, and such particular cases as supercritical carbon dioxide, anthrax lethal factor, supercritical water, vanadium-containing enzymes (relevant to industrial biocatalysis), polymersomes, water clusters, molecular wires, DNA-binding, biological membranes. The properties of interest include molecular conformation, electronic structure, chemical bonding, dynamics, reaction mechanisms, function, transport, diffusion, solvation, intermolecular interactions, hydrogen bonding, ion channel gating, temperature effects, nucleation, electron tunneling, chromatography, electrochemistry, ligand docking, and impact damage. All told it looks to be an extraordinary symposium which is only fitting to honor the remarkable career-to-date of Professor Michael L. Klein. Invited Speakers: Sanjoy Bandyopadhyay, Indian Institute of Technology Paolo Carloni, SISSA/ISAS, Trieste David Chandler, University of California, Berkeley Bin Chen, Louisiana State University David Coker, Boston University Dennis Discher, University of Pennsylvania Carmen Domene, University of Oxford Richard Gee, Lawrence Livermore National Laboratory Francesco Gervasio, ETH (Zurich) Teresa Head-Gordon, University of California, Berkeley Jer-Lai Kuo, Nanyang Technological University Kwang Kim, Pohang University of Science and Technology Kari Laasonen, University of Oulu Uzi Landman, Georgia Institute of Technology Alessandra Magistrato, SISSA/ISAS, Trieste Glen Martyna, IBM Research, TJ Watson Research Center Preston Moore, University of the Sciences in Philadelphia Chris Mundy, Lawrence Livermore National Laboratory Michelle Parrinello, ETH (Zurich) Gren Patey, University of British Columbia Simone Raugei, SISSA/ISAS, Trieste Ursula Rvthlisberger, Federal Institute of Technology EPFL Peter Rossky, University of Texas, Austin Jean-Paul Ryckaert, Universiti Libre de Bruxelles A. Marco Saitta, Universiti Pierre et Marie Curie Giacinto Scoles, Princeton University J. Ilya Siepmann, University of Minnesota John Shelley, Schrvdinger, Inc. Christian Simon, Universiti Pierre et Marie Curie Brian Space, University of South Florida Michiel Sprik, University of Cambridge Yashonath Subramanian, Indian Institute of Science Balasubramian Sundarum, Jawaharlal Nehru Centre for Advanced Scientific Research Mounir Tarek, Universiti Henri Poincari Robert Thomas, University of Oxford Vojislava Pophristic, University of the Sciences in Philadelphia John Tse, University of Saskatchewan Mark Tuckerman, New York University Jost VandeVondele, University of Cambridge Priya Vashista, University of Southern California Symposium webpage: http://www.pharmacy.olemiss.edu/medicinal_chemistry/klein_symposium.html Symposium organizers: University of California, Irvine Chemistry professor Dr. Douglas Tobias and University of Mississippi Medicinal Chemistry professor Dr. Robert Doerksen are co-organizers of the event. They can be reached at: Prof. Doug Tobias Department of Chemistry, University of California at Irvine, Irvine, CA, USA 92697-2025, dtobias-.at.-uci.edu, 949-824-4295 web page: http://chem.ps.uci.edu/research/faculty/dtobias.html Prof. Robert Doerksen Department of Medicinal Chemistry, University of Mississippi, University, MS, USA 38677-1848, rjd-.at.-olemiss.edu, 662-915-5880 web page: http://www.pharmacy.olemiss.edu/medicinal_chemistry/doerksen.html From chemistry-request@ccl.net Tue Feb 8 15:30:52 2005 Received: from relay-1.mail.nethere.net (relay-1.mail.nethere.net [66.63.128.161]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id j18KUni8020496 for ; Tue, 8 Feb 2005 15:30:49 -0500 Received: from scan-1.mail.nethere.net (scan-1.mail [66.63.128.132]) by relay-1.mail.nethere.net (Postfix) with ESMTP id 73AC412BE90; Tue, 8 Feb 2005 10:53:02 -0800 (PST) Received: from mta-2.mail.nethere.net by scan-1.mail.nethere.net with LMTP id 63405-68; Tue, 8 Feb 2005 10:53:02 -0800 (PST) Received: from coffee (unknown [207.167.99.153]) by mta-2.mail.nethere.net (Postfix) with ESMTP id D5D9C87F42; Tue, 8 Feb 2005 10:53:01 -0800 (PST) From: "Ross Walker" To: "'Steve Bowlus'" , "'Rodrigo Pontes'" , Subject: RE: CPU usage Date: Tue, 8 Feb 2005 10:53:01 -0800 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_03AC_01C50DCC.5C102AA0" X-Mailer: Microsoft Office Outlook, Build 11.0.6353 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Thread-Index: AcUNqaBNxxE2j+TsTkCumfPPbOcZqQAZIy4w In-Reply-To: Message-Id: <20050208185301.D5D9C87F42(at)mta-2.mail.nethere.net> X-Spam-Status: No, hits=4.6 required=7.5 tests=HTML_30_40,HTML_FONTCOLOR_BLUE, HTML_MESSAGE,MK_YAHOO_REDIR_01,MY_HTML_OBFU,NO_RDNS,WLS_URI_OPT_150 autolearn=no version=2.61 X-Spam-Checker-Version: SpamAssassin 2.61 (1.212.2.1-2003-12-09-exp) on servernd.ccl.net This is a multi-part message in MIME format. ------=_NextPart_000_03AC_01C50DCC.5C102AA0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit You can lock the processor affinity of a process under windows by running task manager and right clicking on the process and selecting set affinity. You will need to do this with Gaussian's master thread so that all sub threads inherit this affinity. Also, you should double check what you mean by 60% usage. Do you mean 60% of one cpu or 60% of the total? The bar on the left in task manager is the total cpu usage, while the graphs are individual cpu usage. So on a REAL dual cpu machine running one process this bar would show 50% usage. My understanding is that Gaussian under windows is single cpu so this is the best you can hope for. In hyper threaded environments things are a lot more complicated since you don't actually have two cpus, just two pipelines. Hence the total cpu usage gauge can be deceptive... Note, if you manually switch hyper threading off in the bios and still see only 60% cpu utilisation then chances are the process is spending most of its time waiting for memory or swap rather than doing computation. In this case there is not much you can do as you are limited by the machine architecture. So, short answer. "On a hyper threaded machine don't worry about the cpu usage bar it is deceptive.". Only on a real dual cpu machine or single cpu machine is it accurate. All the best Ross /\ \/ |\oss Walker | Department of Molecular Biology TPC15 | | The Scripps Research Institute | | Tel:- +1 858 784 8889 | EMail:- ross(at)rosswalker.co.uk | | http://www.rosswalker.co.uk/ | PGP Key available on request | Note: Electronic Mail is not secure, has no guarantee of delivery, may not be read every day, and should not be used for urgent or sensitive issues. _____ From: Computational Chemistry List [mailto:chemistry-request(at)ccl.net] On Behalf Of Steve Bowlus Sent: 07 February 2005 19:09 To: Rodrigo Pontes; chemistry(at)ccl.net Subject: CCL:CPU usage I have observed the same behavior with another quantum code on a machine with the same CPU/OS combination. There was an earlier post to this list asking about what seemed to be a similar problem with the Xeon HT. I tentatively interpret the behavior that the job is given 100% of one of the HT execution streams, while the other stream is largely kept idle. I have not found a way to tune this behavior. If you look at ControlPanels >AdministrativeTools >ComputerManagement >Processors, you will see there are two CPUs shown. These can be uninstalled, but I haven't tried uninstalling one. Presumably, all CPU resources would be concentrated in a single execution stream, but I haven't found any documentation on this point. I also haven't found anything to suggest the second "CPU" can be _re_installed. I haven't talked to Intel or HP (who built my machine), either of whom could shed light on all these issues. Seems a pity that information on such a feature (or limitation) of HT technology would not be made readily available. sb -----Original Message----- From: Computational Chemistry List [mailto:chemistry-request(at)ccl.net]On Behalf Of Rodrigo Pontes Sent: Monday, February 07, 2005 9:29 AM To: chemistry(at)ccl.net Subject: CCL:CPU usage Dear Colleagues, We are running Gaussian on a Windows XP/Pentiun-IV HT 3.2GHz machine. We observed (in the Task manager / performance) that the CPU usage is only around 60%. Have someone noticed the same problem? How could us set the configuration to a 100% CPU usage? Thanks in advance, Rodrigo M. Pontes Departemanto de Quimica Universidade Estadual de Maringa Brazil _____ Yahoo! Acesso Gratis - Internet rapida e gratis. Instale o discador do Yahoo! agora. ------=_NextPart_000_03AC_01C50DCC.5C102AA0 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

You can lock the processor affinity of a = process under=20 windows by running task manager and right clicking on the process and = selecting=20 set affinity. You will need to do this with Gaussian's master thread so = that all=20 sub threads inherit this affinity.

Also, you should double check what you mean by = 60% usage.=20 Do you mean 60% of one cpu or 60% of the total? The bar on the left in = task=20 manager is the total cpu usage, while the graphs are individual cpu = usage. So on=20 a REAL dual cpu machine running one process this bar would show 50% = usage. My=20 understanding is that Gaussian under windows is single cpu so this is = the best=20 you can hope for. In hyper threaded environments things are a lot = more=20 complicated since you don't actually have two cpus, just two pipelines. = Hence=20 the total cpu usage gauge can be deceptive...

Note, if you manually switch hyper threading = off in the=20 bios and still see only 60% cpu utilisation then chances are the process = is=20 spending most of its time waiting for memory or swap rather than doing=20 computation. In this case there is not much you can do as you are = limited by the=20 machine architecture.

So, short answer. "On a hyper threaded machine = don't worry=20 about the cpu usage bar it is deceptive.". Only on a real dual cpu = machine=20 or single cpu machine is it accurate.

All the best

Ross

From: Computational Chemistry List = [mailto:chemistry-request(at)ccl.net] On Behalf Of Steve=20 Bowlus
Sent: 07 February 2005 19:09
To: Rodrigo = Pontes;=20 chemistry(at)ccl.net
Subject: CCL:CPU = usage

I=20 have observed the same behavior with another quantum code on a machine = with=20 the same CPU/OS combination. There was an earlier post to this = list=20 asking about what seemed to be a similar problem with the Xeon HT.=20

I=20 tentatively interpret the behavior that the job is given 100% of one = of the HT=20 execution streams, while the other stream is largely kept idle. = I have=20 not found a way to tune this behavior. If you look at ControlPanels=20 >AdministrativeTools >ComputerManagement >Processors, = you will=20 see there are two CPUs shown. These can be uninstalled, but I = haven't=20 tried uninstalling one. Presumably, all CPU resources would be=20 concentrated in a single execution stream, but I haven't found any=20 documentation on this point. I also haven't found anything to = suggest=20 the second "CPU" can be _re_installed.

I=20 haven't talked to Intel or HP (who built my machine), either of whom = could=20 shed light on all these issues. Seems a pity that information on = such a=20 feature (or limitation) of HT technology would not be made readily=20 available.

sb

-----Original = Message-----
From:=20 Computational Chemistry List [mailto:chemistry-request(at)ccl.net]On = Behalf Of=20 Rodrigo Pontes
Sent: Monday, February 07, 2005 9:29=20 AM
To: chemistry(at)ccl.net
Subject: CCL:CPU=20 usage

Dear Colleagues,

We are running Gaussian on a Windows XP/Pentiun-IV HT=20 3.2GHz machine. We observed (in the Task manager / performance) = that=20 the CPU usage is only around 60%. Have someone noticed the = same=20 problem? How could us set the configuration to a 100% CPU = usage?

Thanks in advance,

Rodrigo M. Pontes

Departemanto de Quimica

Universidade Estadual de Maringa

Brazil

Yahoo!=20 Acesso Gratis - Internet rapida e gratis. Instale o discador do = Yahoo!=20 agora.

------=_NextPart_000_03AC_01C50DCC.5C102AA0--