Navigation :
Manual
Input Variables
-
Atomic Orbitals
-
Calculation Modes
-
ClassicalParticles
-
DFTBPlusInterface
-
Execution
-
Hamiltonian
--
DFT+U
--
PCM
--
Poisson
---
Multigrid
---
PSolver
---- PoissonSolverPSolverParallelData
--- AlphaFMM
--- DeltaEFMM
--- DressedOrbitals
--- Poisson1DSoftCoulombParam
--- PoissonCutoffRadius
--- PoissonFFTKernel
--- PoissonSolver
--- PoissonSolverBoundaries
--- PoissonSolverMaxIter
--- PoissonSolverMaxMultipole
--- PoissonSolverNodes
--- PoissonSolverThreshold
--- PoissonTestPeriodicThreshold
--
XC
-- AdaptivelyCompressedExchange
-- CalculateSelfInducedMagneticField
-- ClassicalPotential
-- CurrentDensity
-- EnablePhotons
-- EwaldAlpha
-- ExternalCurrent
-- FilterPotentials
-- ForceTotalEnforce
-- GaugeFieldDelay
-- GaugeFieldDynamics
-- GaugeFieldPropagate
-- GaugeVectorField
-- GyromagneticRatio
-- MassScaling
-- MaxwellHamiltonianOperator
-- MaxwellMediumCalculation
-- ParticleMass
-- RashbaSpinOrbitCoupling
-- RelativisticCorrection
-- RiemannSilbersteinSign
-- SOStrength
-- StaticElectricField
-- StaticMagneticField
-- StaticMagneticField2DGauge
-- TheoryLevel
-- TimeZero
-
Linear Response
-
Math
-
Maxwell
-
MaxwellStates
-
Mesh
-
Output
-
SCF
-
States
-
System
-
Time-Dependent
-
Utilities
-
Alphabetic Index
Tutorials
Developers
Releases
PoissonSolverPSolverParallelData
PoissonSolverPSolverParallelData
Section Hamiltonian::Poisson::PSolver
Type logical
Default yes
Indicates whether data is partitioned within the PSolver library.
If data is distributed among processes, Octopus uses parallel data-structures
and, thus, less memory.
If "yes", data is parallelized. The z -axis of the input vector
is split among the MPI processes.
If "no", entire input and output vector is saved in all the MPI processes.
If k-points parallelization is used, "no" must be selected.
Source information
poisson/poisson_psolver.F90 : 193
call parse_variable ( namespace , 'PoissonSolverPSolverParallelData' , . true ., data_is_parallel )