<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <p>Hello David,</p>
    <p>If possible use an optimized blas instead of the reference one,
      as this makes a big difference in performance.</p>
    <p>You might be better off using a serial compiled binary and 2
      threads on each of the 4 images (if with a thread aware
      blas/lapack).<br>
    </p>
    <p>Yes, sorry, had picked up 8 cores and mistakenly thought that was
      the number of images (is 4 enough for REMD?).</p>
    <p>Regards</p>
    <p>Ben<br>
    </p>
    <div class="moz-cite-prefix">On 22/06/2019 17:34, David Furman
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CADtBuo_Ff_cX=Hs-4iiXAjRxyEjbwuHvzSO3_g5aKPei7h40NQ@mail.gmail.com">
      <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div>Hi Ben,</div>
        <div><br>
        </div>
        <div>1. I use libblas-dev 3.6.0 (This  implementation is the
          Fortran 77 reference implementation found  at netlib)</div>
        <div>2. I run each instance from the same shell, 'echo
          $OMP_NUM_THREADS' yields: 1</div>
        <div>3. Yes it is the correct context. I try to run a
          Replica-Exchange MD using 4 replicas. As far as I know, i-pi
          distributes the replicas to clients (DFTB+ in this case) so
          that I can run 4 concurrent replicas. Since I have a
          relatively large system in terms of number of atoms, each
          replica takes a significant amount of time. Hence I was
          thinking of parallelizing each replica on 2 or more MPI cores
          (i.e. parallelize each DFTB+ instance). But, then I encounter
          this weird "bottleneck".</div>
        <div> Regarding your suggestion: I'm not sure how running 8
          serial instances is of help in my case.</div>
        <div><br>
        </div>
        <div>I appreciate your help!</div>
        <div><br>
        </div>
        <div>  <br>
        </div>
      </div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr" class="gmail_attr">On Sat, Jun 22, 2019 at 2:54
          PM <<a
            href="mailto:dftb-plus-user-request@mailman.zfn.uni-bremen.de"
            moz-do-not-send="true">dftb-plus-user-request@mailman.zfn.uni-bremen.de</a>>
          wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0px 0px 0px
          0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Send
          DFTB-Plus-User mailing list submissions to<br>
                  <a
            href="mailto:dftb-plus-user@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user@mailman.zfn.uni-bremen.de</a><br>
          <br>
          To subscribe or unsubscribe via the World Wide Web, visit<br>
                  <a
href="https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user</a><br>
          <br>
          or, via email, send a message with subject or body 'help' to<br>
                  <a
            href="mailto:dftb-plus-user-request@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user-request@mailman.zfn.uni-bremen.de</a><br>
          <br>
          You can reach the person managing the list at<br>
                  <a
            href="mailto:dftb-plus-user-owner@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user-owner@mailman.zfn.uni-bremen.de</a><br>
          <br>
          When replying, please edit your Subject line so it is more
          specific<br>
          than "Re: Contents of DFTB-Plus-User digest..."<br>
          <br>
          <br>
          Today's Topics:<br>
          <br>
             1. Possible bug in DFTB+ with MPI (David Furman)<br>
             2. Re: problem with installation (Farzad Molani)<br>
             3. Re: Possible bug in DFTB+ with MPI (Ben Hourahine)<br>
          <br>
          <br>
----------------------------------------------------------------------<br>
          <br>
          Message: 1<br>
          Date: Sat, 22 Jun 2019 14:41:01 +0100<br>
          From: David Furman <<a href="mailto:sirok4@gmail.com"
            target="_blank" moz-do-not-send="true">sirok4@gmail.com</a>><br>
          To: <a href="mailto:dftb-plus-user@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user@mailman.zfn.uni-bremen.de</a><br>
          Subject: [DFTB-Plus-User] Possible bug in DFTB+ with MPI<br>
          Message-ID:<br>
                  <CADtBuo94A729tJFsTW5ev69sEs4qVNM=<a
            href="mailto:qOXkjMKFTS8_f1N8pg@mail.gmail.com"
            target="_blank" moz-do-not-send="true">qOXkjMKFTS8_f1N8pg@mail.gmail.com</a>><br>
          Content-Type: text/plain; charset="utf-8"<br>
          <br>
           Hi all,<br>
          I have a question concerning the running of DFTB+ with MPI.
          There seems to<br>
          be a problem when running more than one DFTB+ instance (with
          MPI<br>
          parallelization) on the same system.<br>
          <br>
          When I run 4 instances (separate folders) of dftb+ with 2
          cores each (i.e.<br>
          8 cores in total):<br>
          mpirun.openmpi -np 2 dftb+, each %CPU usage drops to 50%.<br>
          <br>
          This is the 'top' output:<br>
          ========================================================<br>
           35426 user    20   0  423384  40404  18248 R  50.5  0.0 
           0:37.69 dftb+<br>
           35434 user    20   0  423080  39720  18248 R  50.2  0.0 
           0:05.70 dftb+<br>
           35396 user    20   0  422044  38660  17644 R  49.8  0.0 
           1:21.35 dftb+<br>
           35402 user    20   0  423384  40460  18304 R  49.8  0.0 
           1:09.54 dftb+<br>
           35403 user    20   0  422044  38872  17848 R  49.8  0.0 
           1:16.15 dftb+<br>
           35427 user    20   0  422044  39048  18032 R  49.8  0.0 
           0:32.93 dftb+<br>
           35435 user    20   0  421732  38164  17716 R  49.8  0.0 
           0:09.07 dftb+<br>
           35395 user    20   0  423384  40360  18204 R  49.5  0.0 
           1:13.18 dftb+<br>
          =========================================================<br>
          <br>
          Whereas, when I run one instance with 8 cores, the efficiency
          is 100% as<br>
          expected:<br>
          <br>
          =========================================================<br>
           35837 user    20   0  441868  35196  19112 R 100.3  0.0 
           0:04.90 dftb+<br>
           35828 user    20   0  442580  35676  19120 R 100.0  0.0 
           0:04.83 dftb+<br>
           35829 user    20   0  442544  36024  19432 R 100.0  0.0 
           0:04.91 dftb+<br>
           35830 user    20   0  441912  35356  19360 R 100.0  0.0 
           0:04.91 dftb+<br>
           35831 user    20   0  441688  35048  19304 R 100.0  0.0 
           0:04.90 dftb+<br>
           35833 user    20   0  441904  35656  19500 R 100.0  0.0 
           0:04.91 dftb+<br>
           35841 user    20   0  441516  34824  19076 R 100.0  0.0 
           0:04.89 dftb+<br>
           35843 user    20   0  441224  34456  18868 R  99.7  0.0 
           0:04.89 dftb+<br>
          =========================================================<br>
          <br>
          I run both cases with OMP_NUM_THREADS=1.<br>
          with open-mpi 1.10.2 and gcc 5.4.0.<br>
          <br>
          Could anyone give a hint about what is wrong?<br>
          -------------- next part --------------<br>
          An HTML attachment was scrubbed...<br>
          URL: <<a
href="http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/96b0d502/attachment-0001.html"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/96b0d502/attachment-0001.html</a>><br>
          <br>
          ------------------------------<br>
          <br>
          Message: 2<br>
          Date: Sat, 22 Jun 2019 21:42:46 +0800<br>
          From: Farzad Molani <<a href="mailto:f.molani@gmail.com"
            target="_blank" moz-do-not-send="true">f.molani@gmail.com</a>><br>
          To: "User list for DFTB+ related questions"<br>
                  <<a
            href="mailto:dftb-plus-user@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user@mailman.zfn.uni-bremen.de</a>><br>
          Subject: Re: [DFTB-Plus-User] problem with installation<br>
          Message-ID:<br>
                  <CAJY0Q+obk-e8iUOJ_986DS4tWpWG4kAZ99a=<a
            href="mailto:6qNstswS0myifg@mail.gmail.com" target="_blank"
            moz-do-not-send="true">6qNstswS0myifg@mail.gmail.com</a>><br>
          Content-Type: text/plain; charset="utf-8"<br>
          <br>
          I know,  but I don't know how can I unstall arpack?<br>
          The package downloaded and now I need help to install it.<br>
          Please note that, I don't permit to login as a root.<br>
          <br>
          On Saturday, June 22, 2019, RICCARDO ROZZA <<a
            href="mailto:rozza.riccardo@studium.unict.it"
            target="_blank" moz-do-not-send="true">rozza.riccardo@studium.unict.it</a>><br>
          wrote:<br>
          <br>
          > Dear Farzad,<br>
          ><br>
          > It seems that you need to install the arpack library (ld:
          cannot find<br>
          > -larpack).<br>
          > All the best,<br>
          ><br>
          > Riccardo.<br>
          ><br>
          > On 22 Jun 2019, at 10:01, Farzad Molani <<a
            href="mailto:f.molani@gmail.com" target="_blank"
            moz-do-not-send="true">f.molani@gmail.com</a>> wrote:<br>
          ><br>
          ><br>
          > Dear DFTB users,<br>
          > I want to install the newest version of DFTB+. The source
          and executable<br>
          > files from <a
            href="http://www.dftbplus.org/download/dftb-stable/"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://www.dftbplus.org/download/dftb-stable/</a> 
          have been<br>
          > downloaded and untarred. I try to install the software
          based on<br>
          > instructions on INSTALL.rst, but I got the following
          error after the make.<br>
          > .<br>
          > .<br>
          > .<br>
          > .<br>
          > w.o dispuffdata.o inputdata.o pmlocalisation.o repcont.o
          repspline.o<br>
          > reppoly.o xlbomd.o extlagrangian.o linresp.o
          linrespgrad.o qm.o<br>
          > linrespcommon.o arpack.o eigensolver.o typegeometry.o
          simplemixer.o<br>
          > numderivs2.o potentials.o thirdorder.o charges.o
          dftbplusu.o geoopt.o<br>
          > steepdesc.o gdiis.o orbitalequiv.o broydenmixer.o
          stress.o etemp.o<br>
          > hermite.o factorial.o intrinsicpr.o mixer.o
          andersonmixer.o<br>
          > elecconstraints.o angmomentum.o eigenvects.o parser.o
          hsdutils2.o<br>
          > unitconversion.o oldcompat.o typegeometryhsd.o
          oldskdata.o densitymatrix.o<br>
          > populations.o repulsive.o spinorbit.o forces.o 
          -L/home/fmolani/DFTB+/<br>
          > dftbplus-18.2/_build/external/xmlf90 -lxmlf90 
          -L/home/fmolani/DFTB+/<br>
          > dftbplus-18.2/_build/external/fsockets -lfsockets<br>
          > -L/apps/intel/2015/composer_xe_2015.3.187/mkl/lib/intel64<br>
          > -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core  -larpack
          -liomp5<br>
          > ld: cannot find -larpack<br>
          > make[1]: *** [dftb+] Error 1<br>
          > make[1]: Leaving directory `/home/fmolani/DFTB+/dftbplus-<br>
          > 18.2/_build/prog/dftb+'<br>
          > make: *** [dftb+] Error 2<br>
          ><br>
          > I would appreciate to help me to solve the problem.<br>
          > Best regards, Farzad<br>
          > --<br>
          > Farzad Molani<br>
          > Postdoctoral Research Fellow<br>
          > School of Pharmaceutical Science and Technology<br>
          > Tianjin University<br>
          > Building 24, Room A205<br>
          > 92 Weijin Road, Nankai District<br>
          > <<a
href="https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g</a>><br>
          > P. R. China<br>
          > <<a
href="https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g</a>><br>
          > Tel.: +8617526526420<br>
          > Email: <a href="mailto:f.molani@gmail.com"
            target="_blank" moz-do-not-send="true">f.molani@gmail.com</a>
          and <a href="mailto:f.molani@tju.edu.cn" target="_blank"
            moz-do-not-send="true">f.molani@tju.edu.cn</a><br>
          > <<a href="mailto:fmolani@dena.kntu.ac.ir"
            target="_blank" moz-do-not-send="true">fmolani@dena.kntu.ac.ir</a>><br>
          ><br>
          ><br>
          > --<br>
          > Farzad Molani<br>
          > Postdoctoral Research Fellow<br>
          > School of Pharmaceutical Science and Technology<br>
          > Tianjin University<br>
          > Building 24, Room A205<br>
          > 92 Weijin Road, Nankai District<br>
          > <<a
href="https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g</a>><br>
          > P. R. China<br>
          > <<a
href="https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://www.google.com/maps/search/92+Weijin+Road,+Nankai+District+%0D%0A+P.+R.+China?entry=gmail&source=g</a>><br>
          > Tel.: +8617526526420<br>
          > Email: <a href="mailto:f.molani@gmail.com"
            target="_blank" moz-do-not-send="true">f.molani@gmail.com</a>
          and <a href="mailto:f.molani@tju.edu.cn" target="_blank"
            moz-do-not-send="true">f.molani@tju.edu.cn</a><br>
          > <<a href="mailto:fmolani@dena.kntu.ac.ir"
            target="_blank" moz-do-not-send="true">fmolani@dena.kntu.ac.ir</a>><br>
          > _______________________________________________<br>
          > DFTB-Plus-User mailing list<br>
          > <a
            href="mailto:DFTB-Plus-User@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">DFTB-Plus-User@mailman.zfn.uni-bremen.de</a><br>
          > <a
href="https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user</a><br>
          ><br>
          ><br>
          ><br>
          <br>
          -- <br>
          Farzad Molani<br>
          Postdoctoral Research Fellow<br>
          School of Pharmaceutical Science and Technology<br>
          Tianjin University<br>
          Building 24, Room A205<br>
          92 Weijin Road, Nankai District<br>
          P. R. China<br>
          Tel.: +8617526526420<br>
          Email: <a href="mailto:f.molani@gmail.com" target="_blank"
            moz-do-not-send="true">f.molani@gmail.com</a> and <a
            href="mailto:f.molani@tju.edu.cn" target="_blank"
            moz-do-not-send="true">f.molani@tju.edu.cn</a> <<a
            href="mailto:fmolani@dena.kntu.ac.ir" target="_blank"
            moz-do-not-send="true">fmolani@dena.kntu.ac.ir</a>><br>
          -------------- next part --------------<br>
          An HTML attachment was scrubbed...<br>
          URL: <<a
href="http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/aae9eaba/attachment-0001.html"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/aae9eaba/attachment-0001.html</a>><br>
          <br>
          ------------------------------<br>
          <br>
          Message: 3<br>
          Date: Sat, 22 Jun 2019 14:54:40 +0100<br>
          From: Ben Hourahine <<a
            href="mailto:benjamin.hourahine@strath.ac.uk"
            target="_blank" moz-do-not-send="true">benjamin.hourahine@strath.ac.uk</a>><br>
          To: <a href="mailto:dftb-plus-user@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">dftb-plus-user@mailman.zfn.uni-bremen.de</a><br>
          Subject: Re: [DFTB-Plus-User] Possible bug in DFTB+ with MPI<br>
          Message-ID: <<a
            href="mailto:16da7e6a-92dd-2184-7a98-067e751cdb14@strath.ac.uk"
            target="_blank" moz-do-not-send="true">16da7e6a-92dd-2184-7a98-067e751cdb14@strath.ac.uk</a>><br>
          Content-Type: text/plain; charset="utf-8"<br>
          <br>
          Hello David,<br>
          <br>
          thoughts so far:<br>
          <br>
          1) Which blas are you using? Sometimes openBlas steals extra
          threads, is<br>
          your top output shown with thread reporting enabled (H)?<br>
          <br>
          2) Are you definitely getting the OMP_NUM_THREADS shell
          variable to each<br>
          of the instances?<br>
          <br>
          3) If this is related to the i-PI mailing list question, is
          there any<br>
          reason not to run 8 serial instances with 1 thread each? The
          MPI<br>
          parallelism overhead on just one process is probably not worth
          it.<br>
          <br>
          Regards<br>
          <br>
          Ben<br>
          <br>
          On 22/06/2019 14:41, David Furman wrote:<br>
          > Hi all,<br>
          > I have a question concerning the running of DFTB+ with
          MPI. There<br>
          > seems to be a problem when running more than one DFTB+
          instance (with<br>
          > MPI parallelization) on the same system.<br>
          ><br>
          > When I run 4 instances (separate folders) of dftb+ with 2
          cores each<br>
          > (i.e. 8 cores in total):<br>
          > mpirun.openmpi -np 2 dftb+, each %CPU usage drops to 50%.<br>
          ><br>
          > This is the 'top' output:<br>
          > ========================================================<br>
          >  35426 user    20   0  423384  40404  18248 R  50.5 
          0.0   0:37.69 dftb+<br>
          >  35434 user    20   0  423080  39720  18248 R  50.2 
          0.0   0:05.70 dftb+<br>
          >  35396 user    20   0  422044  38660  17644 R  49.8 
          0.0   1:21.35 dftb+<br>
          >  35402 user    20   0  423384  40460  18304 R  49.8 
          0.0   1:09.54 dftb+<br>
          >  35403 user    20   0  422044  38872  17848 R  49.8 
          0.0   1:16.15 dftb+<br>
          >  35427 user    20   0  422044  39048  18032 R  49.8 
          0.0   0:32.93 dftb+<br>
          >  35435 user    20   0  421732  38164  17716 R  49.8 
          0.0   0:09.07 dftb+<br>
          >  35395 user    20   0  423384  40360  18204 R  49.5 
          0.0   1:13.18 dftb+<br>
          > =========================================================<br>
          ><br>
          > Whereas, when I run one instance with 8 cores, the
          efficiency is 100%<br>
          > as expected:<br>
          ><br>
          > =========================================================<br>
          >  35837 user    20   0  441868  35196  19112 R 100.3 
          0.0   0:04.90 dftb+<br>
          >  35828 user    20   0  442580  35676  19120 R 100.0 
          0.0   0:04.83 dftb+<br>
          >  35829 user    20   0  442544  36024  19432 R 100.0 
          0.0   0:04.91 dftb+<br>
          >  35830 user    20   0  441912  35356  19360 R 100.0 
          0.0   0:04.91 dftb+<br>
          >  35831 user    20   0  441688  35048  19304 R 100.0 
          0.0   0:04.90 dftb+<br>
          >  35833 user    20   0  441904  35656  19500 R 100.0 
          0.0   0:04.91 dftb+<br>
          >  35841 user    20   0  441516  34824  19076 R 100.0 
          0.0   0:04.89 dftb+<br>
          >  35843 user    20   0  441224  34456  18868 R  99.7 
          0.0   0:04.89 dftb+<br>
          > =========================================================<br>
          ><br>
          > I run both cases with OMP_NUM_THREADS=1.<br>
          > with open-mpi 1.10.2 and gcc 5.4.0.<br>
          ><br>
          > Could anyone give a hint about what is wrong?<br>
          ><br>
          ><br>
          > _______________________________________________<br>
          > DFTB-Plus-User mailing list<br>
          > <a
            href="mailto:DFTB-Plus-User@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">DFTB-Plus-User@mailman.zfn.uni-bremen.de</a><br>
          > <a
href="https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user</a><br>
          <br>
          -- <br>
                  Dr.  B.  Hourahine,  Senior  lecturer<br>
                      SUPA, Department  of  Physics,<br>
                       University  of  Strathclyde,<br>
                        John  Anderson  Building,<br>
                    107 Rottenrow, Glasgow G4 0NG, UK.                 
            <br>
            +44 141 548 2325, <a
            href="mailto:benjamin.hourahine@strath.ac.uk"
            target="_blank" moz-do-not-send="true">benjamin.hourahine@strath.ac.uk</a><br>
          <br>
          The  Department is  a partner  in SUPA,  the Scottish<br>
                      Universities Physics Alliance<br>
          <br>
          The University  of Strathclyde  is a  charitable body,<br>
                 registered in Scotland, number SC015263<br>
          <br>
          -------------- next part --------------<br>
          An HTML attachment was scrubbed...<br>
          URL: <<a
href="http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/bf90b2fb/attachment.html"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20190622/bf90b2fb/attachment.html</a>><br>
          <br>
          ------------------------------<br>
          <br>
          Subject: Digest Footer<br>
          <br>
          _______________________________________________<br>
          DFTB-Plus-User mailing list<br>
          <a href="mailto:DFTB-Plus-User@mailman.zfn.uni-bremen.de"
            target="_blank" moz-do-not-send="true">DFTB-Plus-User@mailman.zfn.uni-bremen.de</a><br>
          <a
href="https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user"
            rel="noreferrer" target="_blank" moz-do-not-send="true">https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user</a><br>
          <br>
          ------------------------------<br>
          <br>
          End of DFTB-Plus-User Digest, Vol 58, Issue 11<br>
          **********************************************<br>
        </blockquote>
      </div>
      <br clear="all">
      <br>
      -- <br>
      <div dir="ltr" class="gmail_signature">
        <div dir="ltr">
          <div>
            <div dir="ltr">
              <pre><span style="font-family:courier new,courier">David Furman, PhD, MRSC<b>
</b>Herchel Smith Research Fellow
Department of Chemistry
University of Cambridge
Lensfield Road
CB2 1EW</span></pre>
            </div>
          </div>
        </div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <pre class="moz-quote-pre" wrap="">_______________________________________________
DFTB-Plus-User mailing list
<a class="moz-txt-link-abbreviated" href="mailto:DFTB-Plus-User@mailman.zfn.uni-bremen.de">DFTB-Plus-User@mailman.zfn.uni-bremen.de</a>
<a class="moz-txt-link-freetext" href="https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user">https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user</a></pre>
    </blockquote>
    <pre class="moz-signature" cols="72">-- 
        Dr.  B.  Hourahine,  Senior  lecturer
            SUPA, Department  of  Physics,
             University  of  Strathclyde,
              John  Anderson  Building,
          107 Rottenrow, Glasgow G4 0NG, UK.                    
  +44 141 548 2325, <a class="moz-txt-link-abbreviated" href="mailto:benjamin.hourahine@strath.ac.uk">benjamin.hourahine@strath.ac.uk</a>

The  Department is  a partner  in SUPA,  the Scottish
            Universities Physics Alliance

The University  of Strathclyde  is a  charitable body,
       registered in Scotland, number SC015263</pre>
  </body>
</html>