Dear Javier, Javier Fuhr wrote: > Hi, > I think I've found a bug in the calculation of the dispersion > energy/force/stress when doing a parallel calculation with more than one > pool. In mm_dispersion.f90, the sum between the processes is only done > intra_pool but not inter_pool.
I don't know that (pretty new) part of the code, but at a first glance to that subroutine I would say that you must be wrong. The parallelization is done inside each pool over atoms, hence you only need to collect the partial sums over subsets of atoms from the processes in the pool. Using more than one pool will not improve the speedup of that part of the code, but neither change the result of energy_london. If you have (reproducible) evidence that the same calculation performed with more than one pool gives a different result, please provide some simple input file. Thanks, GS > Here is a diff between the version 4.1 and a possible corrected one (I'm > not an expert in the paralelization of QE...): > ----------------------------- > 261c261 > < USE mp_global, ONLY : mpime , nproc , intra_pool_comm > --- > > USE mp_global, ONLY : mpime , nproc , intra_pool_comm, > inter_pool_comm > 362a363 > > CALL mp_sum ( energy_london , inter_pool_comm ) > 377c378 > < USE mp_global, ONLY : mpime , nproc , intra_pool_comm > --- > > USE mp_global, ONLY : mpime , nproc , intra_pool_comm, > inter_pool_comm > 503a505 > > CALL mp_sum ( force_london , inter_pool_comm ) > 519c521 > < USE mp_global, ONLY : mpime , nproc , intra_pool_comm > --- > > USE mp_global, ONLY : mpime , nproc , intra_pool_comm, > inter_pool_comm > 660a663 > > CALL mp_sum ( stres_london , inter_pool_comm ) > ----------------------------- > > Regards, > Javier > > -- > Dr. Javier Daniel Fuhr > Grupo F?sica de Superficies - Div. Colisiones At?micas > Centro At?mico Bariloche > Bariloche - Argentina > TE : ++54 2944 44 5100 Int. 5524 > FAX: ++54 2944 44 5299 > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o
