Yambo Community Forum

Hello,

I'm trying to run a calculation for GaAs and I'm getting the following error during the bse run:

<03s> P1: [ERROR] Allocation of GLOBAL_BS_MAT failed with code 5014
P1: [ERROR] STOP signal received while in[02.01] Transition Groups build-up @q1
P1: [ERROR]Attempt to allocate an allocated object

I have generated the screening/dipoles database with no problems, but I can't proceed with the BSE. I'm attaching the log output for one of the mpi during the calculation, and also the bse input and r- output. I'm not sure if it can be memory related, I've been trying to parallelize over k so far. Thank you in advance!
Regards,

Dear Bruno,

this does not seem a memory problem. Not easy to spot the problem.
What I can see is that you provide 8 cpus in the input while running with 64 MPI process. In this case yambo switch to a default assignation, this should not be a problem but can you try to run your job using 8MPI, or more by assigning them consistently and see if the problem disappear.

Best,
Daniele

Hello,

Thank you for the fast reply. Indeed the wrong number of cores on the input was my bad, I was doing some testing and forgot to change it back. Although, using consistent numbers still leads to the same error unfortunately. Not sure of what's happening. I will perform some more tests to investigate.

Regards,

Just a small update. For some reason reducing the number of included bands appears to solve the problem, I ran a quick test with just 1 valence and 1 conduction band and the calculation went through fine. I guess the BSE matrix size was the problem? I will do some more tests.

Regards,