[QE-users] Problem using hse calculation

[Lorenzo Paulatto]

Hello Nawaf,
the HSE calculation has to do a self-consistency on the wavefunctions 
that starts from a standard local-density PBE calculation. Note that the 
HSE calculation, even when using the ACE algorithm, is orders of 
magnitude longer than the PBE one.

It is the PBe calculation that converges in your case, the HSE one is 
too heavy for your computer, it cannot even reach the first step when it 
is killed (after 4 days).

regards


> 
> while trying to do hse calculation ( 47 atoms ) the calculation is not 
> terminating , it converges but ace calculation is not terminating.
> This is how the output file ends
> 
>      the Fermi energy is     7.5626 ev
> 
> !    total energy              =   -7271.48782233 Ry
>       Harris-Foulkes estimate   =   -7271.48803855 Ry
>       estimated scf accuracy    <       0.00023200 Ry
> 
>       convergence has been achieved in  11 iterations
> 
>       Using ACE for calculation of exact exchange
> 
>       EXX grid:    48141 G-vectors     FFT dimensions: (  45,  45,  45)
> Application 969344 exit codes: 1, 139
> Application 969344 resources: utime ~410607s, stime ~626186s, Rss 
> ~43360, inblocks ~1310, outblocks ~0
> -----------------------------------------------------------------------------------------------------
> 
>   18) perftools-base/7.0.0
>   19) PrgEnv-intel/6.0.4
>   20) intel/18.0.1.163 <http://18.0.1.163>
>   21) craype-network-aries
>   22) craype/2.5.13
>   23) cray-mpich/7.7.0
>   24) craype-hugepages2M
>   25) pbs/default
> cray-mpich/7.7.0 requires craype/1.00 or later
> 16.0.4(103):ERROR:105: Unable to locate a modulefile for 
> 'craype-x86-skylake'
> 16.0.4(104):ERROR:105: Unable to locate a modulefile for 
> 'craype-hugepages2M'
> Currently Loaded Modulefiles:
>    1) modules/3.2.10.6 <http://3.2.10.6>
>    2) alps/6.5.30-6.0.5.1_5.1__g60cce32.ari
>    3) nodestat/2.3.78-6.0.5.0_8.13__gbe57af8.ari
>    4) sdb/3.3.751-6.0.5.0_20.25__gb936019.ari
>    5) udreg/2.3.2-6.0.5.0_13.12__ga14955a.ari
>    6) ugni/6.0.14-6.0.5.0_16.9__g19583bb.ari
>    7) gni-headers/5.0.12-6.0.5.0_2.15__g2ef1ebc.ari
>    8) dmapp/7.1.1-6.0.5.0_49.8__g1125556.ari
>    9) xpmem/2.2.4-6.0.5.1_8.6__g35d5e73.ari
>   10) llm/21.3.467-6.0.5.0_25.1__g4c6f300.ari
>   11) nodehealth/5.5.7-6.0.5.0_9.8__g7732908.ari
>   12) system-config/3.5.2545-6.0.5.0_5.1__g8a8b20f.ari
>   13) sysadm/2.4.125-6.0.5.0_16.13__g98e00a9.ari
>   14) Base-opts/2.4.123-6.0.5.0_11.2__g6460790.ari
>   15) rca/2.2.16-6.0.5.0_15.34__g5e09e6d.ari
>   16) craype-network-aries
>   17) craype-hugepages2M
>   18) pbs/default
>   19) PrgEnv-intel/6.0.4
>   20) cray-fftw/3.3.6.2 <http://3.3.6.2>
>   21) intel/16.0.4.258
>   22) /home/apps/qe/6.2.0/modulefiles/int/16.0.4
> =>> PBS: job killed: walltime 345616 exceeded limit 345600
> aprun: Apid 969344: Caught signal Terminated, sending to application
> forrtl: error (78): process killed (SIGTERM)
> Image              PC                Routine            Line        Source
> libintlc.so.5      00002AAAB1EA579D  Unknown               Unknown  Unknown
> libintlc.so.5      00002AAAB1EA3637  Unknown               Unknown  Unknown
> pw.x               0000000020C32DA4  Unknown               Unknown  Unknown
> pw.x               0000000020C32BB6  Unknown               Unknown  Unknown
> pw.x               0000000020B9F659  Unknown               Unknown  Unknown
> pw.x               0000000020BA677E  Unknown               Unknown  Unknown
> Unknown            00002AAAB7168B10  Unknown               Unknown  Unknown
> libpthread.so.0    00002AAAB7165785  Unknown               Unknown  Unknown
> forrtl: error (78): process killed (SIGTERM)
> 
> This is how the file listing modules terminates , the walltime is 96 
> hours. I have attatched my input file . I have tried smaller parameters 
> and smearing instead of tetrahedron , still the same error.
> 
> Thank You.
> Nawaf A
> MSc Energy Science student
> IIT Bombay
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
> 

-- 
Lorenzo Paulatto - Paris