[Smeagol-discuss] Does smeagol require a lof of memory?

XIAOHONG ZHENG exhzheng at gmail.com
Fri May 2 16:13:34 IST 2008


Dear all,

I am doing some calculations for a system in which a C60 cluster is 
sandwiched between two Al electrodes. There are 136 atoms in the 
scattering region. To save memory, SZP or SZ basis is used. However, I 
always got the following error message:

[3]+  Segmentation fault      (core dumped) smeagol < conductor.fdf > 
conductor.out

The last several lines of the output file is:


for SZ basis:

...
initatomlists: Number of atoms, orbitals, and projectors:    136   544  1224

* ProcessorY, Blocksize:    1   8


siesta: System type = chain

* Maximum dynamic memory allocated =     2 MB

siesta:                 ===============================
SMEAGOL                 Bias step =    0, V =  0.000 eV
                            Begin CG move =      0
                        ===============================

InitMesh: MESH =   144 x   144 x   360 =     7464960
InitMesh: Mesh cutoff (required, used) =   200.000   207.901 Ry

* Maximum dynamic memory allocated =   697 MB

For SZP basis:

...
initatomlists: Number of atoms, orbitals, and projectors:    136  1224  2176

* ProcessorY, Blocksize:    1   8


siesta: System type = chain

* Maximum dynamic memory allocated =     2 MB

siesta:                 ===============================
SMEAGOL                 Bias step =    0, V =  0.000 eV
                            Begin CG move =      0
                        ===============================

InitMesh: MESH =   144 x   144 x   360 =     7464960
InitMesh: Mesh cutoff (required, used) =   200.000   207.901 Ry

* Maximum dynamic memory allocated =  1137 MB



This is a serial run. It does not show any other error messages so I do 
not know why it exited abnormally. It should not be due to the 
compilation or library problems since my calculation on a simpler system 
goes on very well. We can see that there is a big memory allocation of 
1137 MB or 697 MB. I have a memory with 2G bytes. Did it exit due to the 
memory problem, please?

Thank you very much!

Sincerely,
Xiaohong




More information about the Smeagol-discuss mailing list