Opened 13 years ago

Closed 12 years ago

Last modified 3 years ago

#77 closed Enhancement (fixed)

LIM3 optimization and parallelisation

Reported by: rblod Owned by: nemo
Priority: normal Milestone:
Component: LIM3 Version: release-2
Severity: Keywords: LIM*
Cc: Review:
MP ready?:
Progress:

Description (last modified by rblod)

Optimization of lim3 in :

  • suppressing useless computation
  • reducing initialization
  • for vectorial computers, add fs_ substitutions
  • suppress or add call to lbclnk

Parallelisation :
For reproducibility, local criterion have to be global, so we have to use mpp_max or mpp_sum even if it is not really necessary.
For ice rheology, only ice processors are working, a sepcail communicator has been created (ncomm_ice) and all reduction operations have now an optional argument corresponding to the communicator.

Commit History (1)

ChangesetAuthorTimeChangeLog
869rblod2008-03-26T10:21:54+01:00

Parallelisation of LIM3. This commit seems to ensure the reproducibility mono/mpp. See ticket #77.

Change History (6)

comment:1 Changed 13 years ago by rblod

  • Description modified (diff)

comment:2 Changed 12 years ago by ctlod

  • Resolution set to fixed
  • Status changed from new to closed

comment:3 Changed 12 years ago by ctlod

  • Milestone set to LIM 3.0

comment:4 Changed 4 years ago by nicolasmartin

  • Keywords LIM added

comment:5 Changed 4 years ago by nicolasmartin

  • Milestone LIM 3.0 deleted

Milestone LIM 3.0 deleted

comment:6 Changed 3 years ago by nemo

  • Keywords LIM* added; LIM removed
Note: See TracTickets for help on using tickets.