Allocation attempt of ... of zero size

Concerns issues with computing quasiparticle corrections to the DFT eigenvalues - i.e., the self-energy within the GW approximation (-g n), or considering the Hartree-Fock exchange only (-x)

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano

Post Reply
dagosta
Posts: 12
Joined: Thu Mar 22, 2018 8:45 am

Allocation attempt of ... of zero size

Post by dagosta » Thu Apr 12, 2018 9:57 am

I am running a GW PPA calculation of a rather simple bulk material. I am getting a series of warnings

[WARNING] [CPU 0] has no poles
[WARNING]Allocation attempt of X_poles_tab of zero size.
[WARNING]Allocation attempt of ordered_grid_index of zero size.
[WARNING]Allocation attempt of coarse_grid_index of zero size.
[WARNING]Allocation attempt of bare_grid_N of zero size.
[WARNING]Allocation attempt of coarse_grid_Pt of zero size.
[WARNING]Allocation attempt of poles of zero size.
[WARNING]Allocation attempt of Ein_poles of zero size.
[WARNING]Allocation attempt of poles of zero size.
[WARNING]Allocation attempt of Ein_poles of zero size.
[WARNING]Allocation attempt of X_poles of zero size.
[WARNING]Allocation attempt of X_Ein_poles of zero size.
<snip>

and then at the end of the calculation a collection of NaN

What could be the origin of this problem? I tried starting from scratch, i.e., from the QE calculations but I hit the same wall again.

Thanks,
Roberto
Roberto D'Agosta
Nano-Bio Spectroscopy Group
Av de Tolosa 72
Donostia-San Sebastian
Spain

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Allocation attempt of ... of zero size

Post by Daniele Varsano » Thu Apr 12, 2018 10:07 am

Ciao Roberto,
can you upload the input and report file? I suspect there is some variable not set correctly.
Best,

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

dagosta
Posts: 12
Joined: Thu Mar 22, 2018 8:45 am

Re: Allocation attempt of ... of zero size

Post by dagosta » Thu Apr 12, 2018 11:45 am

I am not sure I am authorised to upload attachments...

Roberto
Roberto D'Agosta
Nano-Bio Spectroscopy Group
Av de Tolosa 72
Donostia-San Sebastian
Spain

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Allocation attempt of ... of zero size

Post by Daniele Varsano » Thu Apr 12, 2018 1:02 pm

Dear Roberto,
I'm not sure you sent me the correct input file as in the report it is indicated as gw_ppa.in, probably you just renamed it.
A possible problem could be that something failed in the default parallelization. I can't see how the parallelization has been set as
it is reported in the log file only, as it looks the cpu1 has zero workload.
You can try to assign a parallelization strategy e.g. using 60 cpus as it seems you are using:

Code: Select all

X_all_q_CPU= "1 2 15 2"                # [PARALLEL] CPUs for each role
X_all_q_ROLEs= "q k c v"              # [PARALLEL] CPUs roles (q,k,c,v)
X_all_q_nCPU_LinAlg_INV= 1     # [PARALLEL] CPUs for Linear Algebra
X_Threads=1                  # [OPENMP/X] Number of threads for response functions
DIP_Threads=1                  # [OPENMP/X] Number of threads for dipoles
SE_CPU= "1 2 30"                     # [PARALLEL] CPUs for each role
SE_ROLEs= "q qp b"                   # [PARALLEL] CPUs roles (q,qp,b)
SE_Threads=1   
and see if the problem persists.
Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

dagosta
Posts: 12
Joined: Thu Mar 22, 2018 8:45 am

Re: Allocation attempt of ... of zero size

Post by dagosta » Thu Apr 12, 2018 1:13 pm

Dear Daniele,

Thanks I will check your recommendation and see what I get.

Best regards and sorry for the problem with the attachment.

Roberto
Roberto D'Agosta
Nano-Bio Spectroscopy Group
Av de Tolosa 72
Donostia-San Sebastian
Spain

Post Reply