> hi dear all, > > i am doing a FC run over a system with 189 atoms in the unit cell (formed > by > making use of Supercell). > > but i got the error.....probably bcoz of memory?
Hi, I don't know about your segmentation fault, but the first error message you got is in fact not an error, but a feature: *Cannot do O(N) with k-points.* As you generated a supercell with > 100 atoms, the SolutionMethod turned, by default, from Diagon to OrderN. If OrderN is not an option you absolutely want to use (and as you are doing a crystal not a molecule it could be tricky), set SolutionMethod to Diagon explicitly. Another observation: do you need your supercell really so elongated? It might be not so optimal in view of making real-space grid integration. (Although, with sufficiently high MeshCutoff this shouldn't make a difference...) Good luck Andrei Postnikov > > can anybody help me.? > > last part of my output file is: > > siesta: ============================== > Begin FC step = 0 > Undisplaced coordinates > ============================== > > superc: Internal auxiliary supercell: 2 x 2 x 2 = 8 > superc: Number of atoms, orbitals, and projectors: 1512 17928 22680 > > outcell: Unit cell vectors (Ang): > 15.510000 0.000000 0.000000 > 10.062640 11.802645 0.000000 > 10.062640 4.643694 10.850486 > > outcell: Cell vector modules (Ang) : 15.510000 15.509969 15.509792 > outcell: Cell angles (23,13,12) (deg): 49.5514 49.5494 49.5499 > outcell: Cell volume (Ang**3) : 1986.2793 > > InitMesh: MESH = 192 x 192 x 192 = 7077888 > InitMesh: Mesh cutoff (required, used) = 500.000 553.018 Ry > > * Maximum dynamic memory allocated = 880 MB > *Cannot do O(N) with k-points.* > Stopping Program from Node: 0 > forrtl: severe (174): SIGSEGV,* segmentation fault occurred*
