issues for (nqf1, nqf2, nqf3) > (90,90,1)

[eliephys78]

Dear all,

I am trying to test the convergence of lambda w.r.t to the fine phonon grid. I have set my electronic grid to nkf1=90, nqf2=90, nqf3=1. When I try to start the calculations for a fine q grid higher than 90x 90 x 1, the code stops with a lot of "btr" files (I have no idea what these are!) and an error:

epw.x:24352 terminated with signal 11 at PC=ce14d7 SP=7fff7de70720. Backtrace:

epw.x:24353 terminated with signal 11 at PC=ce14d7 SP=7fff763d96a0. Backtrace:

epw.x:12895 terminated with signal 11 at PC=ce14d7 SP=7fff65650920. Backtrace:

.
.
.
.
(the list goes on)

Whta could the possile source of this error be?

Regards

[eliephys78]

It seems it worked when both nqf and nkf grids were set to 100 x 100 x 1 . The error simply disappeared!

Weird

[sponce]

Hello eliephys78,

Weird indeed. Probably a memory issue or instability on your cluster.

If you re-submit now a calculation with 90x90x1, does it still appears ?

Cheers,
Samuel