ABINIT Discussion Forums

Hello,

I am trying to perform solve for the MEP of a particular system, using the NEB method under the improved tangent implementation. I started with just 6 images, keeping the first and last fixed, so far the code seems to be working, or at least I have not received an error message yet, however I think my resources request is a little conservative.

This is my input file:

#Definition of parallelism
autoparal 1

#Definition of GGA-PBE functional
ixc 11

#Definition of +U variables
usepawu 1 #Full Localized Limit Method
lpawu 3 #Alpha-Uranium has f-orbital occupation
upawu 1.24 eV #From Wei Xei et al
#jpawu 0 #Let abinit deal with default values for J

#Definition of spin
occopt 4 #Alpha-U has metallic occupation
nsppol 2 #occopt is 4, metallic occupation
nspinor 1 #Tied up with nsppol 2
nspden 2 #Sacalar magnetization

#Definition of the SCF cycle
toldff 1.0d-3 #Use force difference as stopping creteria
nstep 100 #Max. Number of SCF steps
diemac 1.0d6 #Value recommended for metals

#Definition of the unit cell
rprim 1.000000000 0.000000000 0.000000000
0.000000000 1.000000000 0.000000000
0.000000000 0.000000000 1.000000000

#Atomic data
natom 31 #There are four atoms within the conventional cell
nband 651 #14 Bands per U atom, 50%
ntypat 1 #There is only one type of atom
znucl 92 #Z number for U
typat 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 #We are dealing with just one type of atom

#Definition of the k-point grid
#kptopt 1
#nshiftk 1
#shiftk 0.0 0.0 0.0
#ngkpt 2 2 2
kptopt 0
nkpt 1
kpt 0 0 0 #Gamma point

#Definition of the path
acell 10.363267115 22.322895979 18.772253591
xred -0.000577159 -0.000084567 -0.000850529
0.250000000 0.249223904 -0.001104593
-0.000891373 0.100497575 0.250000000
0.250000000 0.350896052 0.250000000
0.500577159 -0.000084567 -0.000850529
0.750000000 0.250430685 -0.000356463
0.500891373 0.100497575 0.250000000
0.750000000 0.350865781 0.250000000
-0.000472425 0.500313899 0.000119583
0.250000000 0.751403498 -0.004526546
0.000677969 0.600673331 0.250000000
0.250000000 0.851189815 0.250000000
0.500472425 0.500313899 0.000119583
0.750000000 0.754076168 -0.003584907
0.499322031 0.600673331 0.250000000
0.750000000 0.849381702 0.250000000
-0.000577159 -0.000084567 0.500850529
0.250000000 0.249223904 0.501104593
-0.002337565 0.098849122 0.750000000
0.250000000 0.350820732 0.750000000
0.500577159 -0.000084567 0.500850529
0.750000000 0.250430685 0.500356463
0.502337565 0.098849122 0.750000000
0.750000000 0.350134097 0.750000000
-0.000472425 0.500313899 0.499880417
0.250000000 0.751403498 0.504526546
-0.000131717 0.601291299 0.750000000
0.250000000 0.847521076 0.750000000
0.500472425 0.500313899 0.499880417
0.750000000 0.754076168 0.503584907
0.500131717 0.601291299 0.750000000

acell_lastimg 10.354945024 22.284282376 18.744231480
xred_lastimg 0.000303843 0.000167714 -0.000175325
0.250048090 0.250567720 0.000272182
0.000505010 0.100548368 0.249838112
0.249961763 0.350599803 0.249831300
0.499675029 0.000214860 -0.000367106
0.749959149 0.249504679 -0.000733663
0.499494711 0.100657230 0.249858875
0.750027720 0.350795457 0.250231967
0.000153205 0.500530048 0.000179132
0.249968011 0.752981726 -0.003637382
-0.000717542 0.600233275 0.250212852
0.249958834 0.849762134 0.250266787
0.499850247 0.500509751 0.000275914
0.750037718 0.751368743 -0.003455412
0.500710811 0.600300384 0.249746875
0.750035969 0.850697224 0.249455367
0.000386601 0.000228154 0.500172179
0.249912048 0.250549252 0.499727314
0.001806549 0.099157349 0.750165761
0.249978123 0.350306177 0.750132560
0.499615972 0.000275965 0.500354226
0.750080011 0.249582161 0.500709735
0.498224895 0.099240700 0.750128481
0.749996353 0.350895902 0.749746298
0.000260724 0.500520503 0.499788643
0.249893080 0.752900783 0.503510226
0.000007706 0.601106496 0.749781602
0.499729428 0.500409044 0.499703914
0.750155381 0.751221296 0.503458933
0.499945965 0.601218731 0.750237098
0.750034594 0.847566116 0.750582555

nimage 6 #Number of replicas of the system
imgmov 5 #Nudged Elastic Band Method
neb_algo 1 #Improved tangent option
ntimimage 20 #Max. Number of relaxation steps
tolimg 1.0d-3 #Stopping criteria for image relaxation
dynimage 0 4*1 0 #Keep initial and final states fixed
mep_solver 2 #L-BFGS Algorithm, not sure if this is redundant with SCF definition
prtvolimg 0 #Complete output file

#General input data
ecut 24 #From convergence studies
pawecutdg 50 #From convergence studies
optforces 1 #Enforce calculation of forces at each SCF step
chksymbreak 0 #To by-pass the warning message on the value of tnons
chkprim 0 #Do not check if cell is primitive
nsym 1 #Turn off symmetry

#End of input file

So far, I have several output files, but no numerical results yet (the job has been running for 11 hours at the date of this message). After checking the output file, I noticed that the number of processors differ from what I requested (336 versus 384), and perhaps this is linked to the "autoparal" tool. Could you please give me a hand with this? I think I need to request more processors, especially given the size of my system, but I am not sure how the number of processors per image is assigned.

Thank you,
Manuel

Greetings all,

Has anyone experienced the issue described below?

forrtl: severe (174): SIGSEGV, segmentation fault occurred

Thank you,
Manuel