[Pw_forum] pw.x: start MD with defined velocity

2013-06-26 Thread Steve Schmerler
On Jun 26 16:03 +0200, Steve Schmerler wrote: > Hello > > I have a MD run which I want to repeat from some point, using the same > velocities from that time step. pw.x doesn't have an input card > ATOMIC_VELOCITIES like cp.x, so I need to perform a restart. This is not docume

[Pw_forum] pw.x: start MD with defined velocity

2013-06-26 Thread Steve Schmerler
hem or both for a restart? Thank you. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Freiberg, Germany

[Pw_forum] unit of mode vectors in matdyn.modes

2012-12-13 Thread Steve Schmerler
, obviously including the mass factor. Also routine dyndiag() says that "displacements are eigenvectors divided by sqrt(amass)". Thank you very much. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] unit of mode vectors in matdyn.modes

2012-12-13 Thread Steve Schmerler
00243 Of course! Thank you. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] unit of mode vectors in matdyn.modes

2012-12-13 Thread Steve Schmerler
re out the unit / normalization of * the eigenvectors in matdyn.modes * the q-points (I guess fractional w.r.t. reciprocal unit cell) Thank you. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] parallel phonon calculation on distributed memory machines

2012-10-12 Thread Steve Schmerler
upported as of 4.3.x . best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] parallel phonon calculation on distributed memory machines

2012-10-10 Thread Steve Schmerler
/_ph0/prefix.dwf1 ... /outdir/_ph0/prefix.ebar1 ... /outdir/_ph0/prefix.phsave/ /outdir/_ph0/prefix.save/ This is with QE 4.3.2. Thanks. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] parallel phonon calculation on distributed memory machines

2012-10-10 Thread Steve Schmerler
which runs with 4.2.1, but not with 4.3.2. Havent tested 5.0.1 yet. So what's the situation? Is it now generally recomended to split q-points (and irreps) and run each on a SMP machine? Thank you. best, Steve -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipzig

[Pw_forum] cluster*.in tests fail with PAW and ifort -xHost

2010-12-22 Thread Steve Schmerler
uld I recompile? Thanks. best, Steve [1] http://qe-forge.org/frs/download.php/115/espresso-4.2.1-examples.tar.gz -- Steve Schmerler Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany

[Pw_forum] nstep not honored in vc-md

2010-12-08 Thread Steve Schmerler
On Dec 08 14:36 +0100, Paolo Giannozzi wrote: > did you use "Reply" to a previous message and edited > the subject? No, that's why I wondered. As far as I can see, there are no In-Reply-To, References (or some such) headers in the message. best, Steve -- Steve S

[Pw_forum] nstep not honored in vc-md

2010-12-08 Thread Steve Schmerler
On Dec 08 14:07 +0100, Paolo Giannozzi wrote: > > On Dec 7, 2010, at 18:13 , Steve Schmerler wrote: > > > I'm trying some basic variable cell MD (pw.x, vc-md) and I found that > > the code does not stop after 'nstep' steps with cell_dynamics='w' or &

[Pw_forum] nstep not honored in vc-md

2010-12-07 Thread Steve Schmerler
ile it does so with normal 'md' and 'vc-relax' (cell_dynamics='damp-w'). This is with 4.2.1 and 4.1.2 . Attached are input (pw.in) and output (pw.out, killed after 9 steps, nstep=3 in pw.in). "nstep=3" is mentioned correctly in the output, though. Thank you