Changes between Version 19 and Version 20 of 2013WP/2013Action_institutions_CMCC

Timestamp:

2013-10-30T08:55:15+01:00 (10 years ago)

Author:

vichi

Comment:

--

2013WP/2013Action_institutions_CMCC

@@ -27 +27 @@
 || Action || Brief description || Status || Branch name || Trac ticket n° || Wiki page name || Reviewer(s) || Review status || Nb of weeks || Ready for merge/ If "NO", reason and status for 2014 ||
 || CMCC-1a || Scalability || Continuation || https://forge.ipsl.jussieu.fr/nemo/browser/branches/2013/dev_r3948_CMCC_NorthFold_Opt || https://forge.ipsl.jussieu.fr/nemo/ticket/1150 || || A. Coward || DONE || || YES ||
-|| CMCC-1b || Scalability: MPI-OPENMP || Continuation || [[BR]]https://forge.ipsl.jussieu.fr/nemo/browser/branches/2013/dev_r4017_CMCC_MPI_OpenMP || [[BR]]https://forge.ipsl.jussieu.fr/nemo/ticket/1151 || || || || || NO[[BR]]This is a long term activity. Currently only the GYRE conf is fully parallelized with OpenMP+MPI.[[BR]][[BR]]'''See further notes at CMCC.1 below''' ||
+|| CMCC-1b || Scalability: MPI-OPENMP || Continuation || [[BR]]https://forge.ipsl.jussieu.fr/nemo/browser/branches/2013/dev_r4017_CMCC_MPI_OpenMP || [[BR]]https://forge.ipsl.jussieu.fr/nemo/ticket/1151 || || || || || NO[[BR]]The hybrid OpenMP/MPI approach has been implemented for the Gyre configuration and scalability studies are on going. The development will be continued in 2014[[BR]][[BR]]'''See further notes at CMCC.1 below''' ||
 || CMCC-2 || Masks || Started || || || || || || || ||
 || CMCC-3 || Data assimilation || Not Started || || || || || || || ||
@@ -36 +36 @@
 
 === '''CMCC.1''' Scalability Optimization ===
-'''          Motivation:''' Improve NEMO scalability. [[BR]]  '''          Status:''' Partly closed. The feasibility study related to the introduction of the PETSc library  has been concluded. From this study it has been deduced that a  consistent manpower is needed. Since this effort is not available using  the currently funded projects, the activity related to the introduction  of PETSc into the NEMO code will be sustained when new project funds  will be achieved. A report on the results of this study is  available and will be provided at the November merge party.[[BR]]  '''          Main tasks:'''[[BR]]  * Introduction of a new software layer in the NEMO stack software exploiting numerical optimized and parallel libraries such as  pBLAS, ScaLAPACK and PETSc. [[BR]]  * Continuation of the implementation of hybrid parallel approach based on the functional parallelism for the tracers, momentum equations and 3D domain decomposition.[[BR]]  * Test performances of XIOS on CMCC architectures[[BR]] '''System Reviewer:''' [[BR]]  '''          Deadline:'''  2013[[BR]]  '''          Priority: '''Medium [[BR]]  '''          Principal investigator:''' Italo Epicoco [[BR]]
+'''          Motivation:''' Improve NEMO scalability. [[BR]]  '''          Status:''' Partly closed. The feasibility study related to the introduction of the PETSc library  has been concluded. From this study it has been deduced that a  consistent manpower is needed. Since this effort is not available using  the currently funded projects, the activity related to the introduction  of PETSc into the NEMO code will be sustained when new project funds  will be achieved. A report on the results of this study is  available and will be provided at the November merge party.[[BR]]  '''          Main tasks:'''[[BR]]  * Introduction of a new software layer in the NEMO stack software exploiting numerical optimized and parallel libraries such as  pBLAS, ScaLAPACK and PETSc. [[BR]]  * Continuation of the implementation of hybrid parallel approach based on the functional parallelism for the tracers, momentum equations and 3D domain decomposition.[[BR]]  * Test performances of XIOS on CMCC architectures[[BR]] '''System Reviewer:''' [[BR]]  '''          Deadline:'''  2013[[BR]]  '''          Priority: '''Medium [[BR]]  '''          Principal investigator:''' Italo Epicoco [[BR]]
 
 === '''CMCC.2 '''         Modification of masks ===