geometry optimization unstable

sandeep kuma... at gmail.com
Mon Aug 6 03:12:18 UTC 2012
Previous message (by thread): CELL_OPT keyword
Next message (by thread): geometry optimization unstable
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Dear all,
                I am doing geometry optimization of a system containing 576 
atoms. When i ran it a machine having intel X5460 processor, i didn't see 
any oscillations in total energies. It converges within some steps. But, 
the same system shows oscillations on different machine (intel X5670 
processor). 

*Total energy :
*

 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857268930485588
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857593977155830
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.856003472735210
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857548468864024
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.854062773488295
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857594769232037
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.854626921172439
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857481194438606
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.852779045098487
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.857273394239655
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.851571809396773
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.856224749560624
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.846230635967004
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.841851463337662
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.847479397051302
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.831469604048834
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.842533189635105
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.826210614017327
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.837802791426839
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.809268982869980
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.832563770153229
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.794880845741773
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.819193957763218
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.745714916703946
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.820245638992219
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.648271216216926
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.811590467374117
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.573351584592274
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.807336936256434
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.548983575820785
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.804035882624248
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.454113226169284
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.798586604405500
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.406644262456666
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.795520566227879
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.400278104366862
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.804283443121221
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.585113812539021
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.813855948710625
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.532866042457499
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.817657287906968
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.571135385605430
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.819301508552599
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.565533391461031
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.817356171838128
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.594185149152509
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.816696580745429
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.537045230232707
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.813409116061393
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.549188327705451
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.806053104601233
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.659599977649577
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.810634872865649
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.540129332885954
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.808556044454690
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.522990522475538
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.821163895395784
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.541664381568353
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.811375715670692
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.514473791798991
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.801293038222866
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.521168700564431
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.806579234558285
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.456123853706231
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.811449793531210
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.576983119674424
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.816395686705164
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.550635824096389
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.815701719725894
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.489804006631857
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.819771494755969
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.566217729472100
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.802585890487080
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.508774924785939
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.800921016420943
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.544680224335480
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.822501357248711
 ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
 -3895.651280964678790


*Makefile :*


INTEL_INC=/opt/intel/mkl/10.2.0.013/include
INTEL_LIB=/opt/intel/mkl/10.2.0.013/lib/em64t/
FFT_INC=/opt/intel/mkl/10.2.0.013/include/fftw


CC       = cc
CPP      = 
FC       = mpiifort
LD       = mpiifort
AR       = ar -r
DFLAGS   = -D__INTEL -D__FFTSG -D__parallel -D__BLACS -D__SCALAPACK 
-D__FFTW3
CPPFLAGS = 
FCFLAGS  = $(DFLAGS) -I$(INTEL_INC) -I$(FFT_INC) 
-I/opt/intel/impi/4.0.2/include64  -O3 -msse2 -heap-arrays 64 
-funroll-loops -fpp -free
FCFLAGS2 = $(DFLAGS) -I$(INTEL_INC) -I$(FFT_INC) -O1 -msse2 -heap-arrays 64 
-fpp -free
LDFLAGS  = $(FCFLAGS) -I$(INTEL_INC)
LIBS =  -L$(MKLROOT)/lib/intel64 
 $(MKLROOT)/lib/intel64/libmkl_blas95_lp64.a 
 $(MKLROOT)/lib/intel64/libmkl_lapack95_lp64.a \
 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core 
-lmkl_blacs_intelmpi_lp64 -lpthread -lm $(INTEL_LIB)/libfftw3xf_intel.a 


OBJECTS_ARCHITECTURE = machine_intel.o


graphcon.o: graphcon.F
$(FC) -c $(FCFLAGS2) $<

--------------------------------------------------------------------------

I  used same Makefile on both architectures (Intel compilers 12.1).  I 
tried with different optimization levels, but this too didn't help. And 
also, it's not scaling more than one node on this machine. (each node 
contains 12 processors).

Do anybody faced similar problem ?  Any help in this regard is greatly 
appreciated.

Regards,
Sandeep





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20120805/17e06d53/attachment.htm>
Previous message (by thread): CELL_OPT keyword
Next message (by thread): geometry optimization unstable
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the CP2K-user mailing list