Mem All. failed. Element WF require

[Stephan]

Hi,
I try to simulate the optical response of BaFe2As2 using bse.
This is the input file:

#
# ::: ::: ::: :::: :::: ::::::::: ::::::::
# :+: :+: :+: :+: +:+:+: :+:+:+ :+: :+: :+: :+:
# +:+ +:+ +:+ +:+ +:+ +:+:+ +:+ +:+ +:+ +:+ +:+
# +#++: +#++:++#++: +#+ +:+ +#+ +#++:++#+ +#+ +:+
# +#+ +#+ +#+ +#+ +#+ +#+ +#+ +#+ +#+
# #+# #+# #+# #+# #+# #+# #+# #+# #+#
# ### ### ### ### ### ######### ########
#
# GPL Version 3.4.1 Revision 3187
# http://www.yambo-code.org
#
optics # [R OPT] Optics
bse # [R BSE] Bethe Salpeter Equation.
BSEmod= "causal" # [BSE] resonant/causal/coupling
BSKmod= "Hartree" # [BSE] IP/Hartree/HF/ALDA/SEX
% QpntsRXd
1 | 1 | # [Xd] Transferred momenta
%
% BEnRange
0.00000 | 3.50000 | eV # [BSS] Energy range
%
% BDmRange
0.10000 | 0.10000 | eV # [BSS] Damping range
%
BEnSteps= 100 # [BSS] Energy steps
% BLongDir
1.000000 | 0.000000 | 0.000000 | # [BSS] [cc] Electric Field
%
% BSEBands
1 | 80 | # [BSK] Bands range
%
DrudeWBS= (2.54169,0.371955) eV # [Xd] Drude plasmon
%

But at the end of the simulation I received the error message
ERROR] STOP signal received while in :[04] Bethe-Salpeter Kernel
[ERROR]Mem All. failed. Element WF require -13.06491 [Gb]

There fore I cleaned my memory a little bit and gained about 100GB more space.
But still I receive the error message
ERROR] STOP signal received while in :[04] Bethe-Salpeter Kernel
[ERROR]Mem All. failed. Element WF require -13.06491 [Gb]

I thougt that means I have 13.06491 Gb too less. Am I wrong...?

The whole output in terminal:

pi1@QuanCal:~/Desktop/FinalsCa/CaFe2As2/297K/nscf_31.3/CaFe2As2_297K.save$ /home/pi1/Desktop/Simulation/yambo-3.4.1-rev61/bin/yambo

____ ____ ___ .___ ___. .______ ______
\ \ / / / \ | \/ | | _ \ / __ \
\ \/ / / ^ \ | \ / | | |_) | | | | |
\_ _/ / /_\ \ | |\/| | | _ < | | | |
| | / _____ \ | | | | | |_) | | `--" |
|__| /__/ \__\ |__| |__| |______/ \______/

<---> [01] Files & I/O Directories
<---> [02] CORE Variables Setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [WARNING]Metallic system
<---> [03] Transferred momenta grid
<---> [M 0.747 Gb] Alloc bare_qpg ( 0.581)
<04s> [03.01] Main loop
<05s> [X/K] Drude contribution @[ev]: 2.541690 0.371955
<05s> [X/K] Drude reference freq. for w=0 is [eV]: 0.000000
<16s> [04] Bethe-Salpeter Kernel
<16s> [BSE] Kernel dimension : 4994311
<02h-53m-35s> Filling Estimation |# | [005%] 02h-53m-18s(E) 02d-09h-45m- <05h-14m-24s> Filling Estimation |## | [010%] 05h-14m-07s(E) 02d-04h-20m- <08h-18m-02s> Filling Estimation |### | [015%] 08h-17m-45s(E) 02d-07h-18m- <11h-17m-55s> Filling Estimation |#### | [020%] 11h-17m-38s(E) 02d-08h-27m- <14h-26m-23s> Filling Estimation |##### | [025%] 14h-26m-06s(E) 02d-09h-44m- <17h-18m-23s> Filling Estimation |###### | [030%] 17h-18m-06s(E) 02d-09h-40m- <19h-11m-24s> Filling Estimation |####### | [035%] 19h-11m-07s(E) 02d-06h-48m- <01d-01h-25m-07s> Filling Estimation |######### | [045%] 01d-01h-24m-50s(E) 02d-08h-28m-24 <01d-04h-02m-50s> Filling Estimation |########## | [050%] 01d-04h-02m-34s(E) 02d-08h-05m-08 <01d-07h-07m-49s> Filling Estimation |########### | [055%] 01d-07h-07m-32s(E) 02d-08h-35m-29 <01d-08h-46m-54s> Filling Estimation |############ | [060%] 01d-08h-46m-37s(E) 02d-06h-37m-39 <01d-11h-37m-36s> Filling Estimation |############# | [065%] 01d-11h-37m-19s(E) 02d-06h-48m-06 <01d-14h-37m-44s> Filling Estimation |############## | [070%] 01d-14h-37m-27s(E) 02d-07h-10m-34 <01d-16h-32m-54s> Filling Estimation |############### | [075%] 01d-16h-32m-38s(E) 02d-06h-03m-30 <01d-18h-18m-50s> Filling Estimation |################ | [080%] 01d-18h-18m-33s(E) 02d-04h-53m-10 <01d-21h-16m-10s> Filling Estimation |################# | [085%] 01d-21h-15m-54s(E) 02d-05h-15m-08 <01d-22h-49m-42s> Filling Estimation |################## | [090%] 01d-22h-49m-25s(E) 02d-04h-01m-31 <02d-00h-47m-25s> Filling Estimation |################### | [095%] 02d-00h-47m-08s(E) 02d-03h-21m-08 <02d-02h-01m-37s> Filling Estimation |####################| [100%] 02d-02h-01m-20s(E) 02d-02h-01m-20s(X)
<02d-02h-01m-37s> [P 01] Kernel filling [o/o] 100.0000

[ERROR] STOP signal received while in :[04] Bethe-Salpeter Kernel
[ERROR]Mem All. failed. Element WF require -13.06491 [Gb]



Thanks and regards

Stephan

[Daniele Varsano]

Dear Stephan,
please note that the mem failing refers to the RAM memory and not the HD memory.
First of all, please note that you have an enormous excitonic matrix> dimension 4994311.
Are you sure you need all that 80 bands? Not having posted the report file I cannot see the k point sampling.
I would start to reduce the matrix choosing few bans below and above the metallic bands and then add bands until convergences.
Surely you can skip the first bands that should be very deep in energy.
The dimension of your matrix looks to me crazy.
Next you can try to reduce the memory required by the wfs by using the FFTGvecs input variable.
Check in your report file , r_setup , the number of G vector appearing in the ns.db1 database, and reduce that number (again without the report I cannot suggest you any values, but check the WFs gvectors and reduce it a bit (not too much otherwise you will loose otrhonormality).

Hope it helps,
Daniele