VISCOELASTIC_LOAD and VISCOELASTIC_LOAD_PML are currently implemented, but VISCOACOUSTIC_LOAD and VISCOACOUSTIC_LOAD_PML are not.
This means that viscoacoustic elements (with or without PML) are currently much slower than they could in the case of MPI runs because they currently use the non-dissipative cost (weight) in domain decomposition with SCOTCH.
VISCOELASTIC_LOAD and VISCOELASTIC_LOAD_PML are currently implemented, but VISCOACOUSTIC_LOAD and VISCOACOUSTIC_LOAD_PML are not.
This means that viscoacoustic elements (with or without PML) are currently much slower than they could in the case of MPI runs because they currently use the non-dissipative cost (weight) in domain decomposition with SCOTCH.
Serial runs (no MPI) are not affected.
Dimitri.