Changes between Version 3 and Version 4 of ticket/0829


Ignore:
Timestamp:
2011-05-31T11:06:15+02:00 (9 years ago)
Author:
rblod
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • ticket/0829

    v3 v4  
    2222 * More radical change : the working space is build dynamically. This solves limitation 2. 
    2323 
    24 ==== Timing ==== 
    25 ==== Auto-assignement ==== 
    26 ==== Dynamic dynamic memory ==== 
     24==== 1- Timing ==== 
     25 
     26This functionality doesn't aim to replace advanced software used for optimisation but: 
     27 * to give a rough idea of performance (CPU and elapsed) 
     28 * use the same tools and format on all computers( fortran intrinsec CPU_TIME and WMPI_TIME) 
     29It is bases on a linked chain of informations to be able to add dynamically new sections and add sub-sections[[BR]] 
     30Implementation: 
     31 * CALL timing_init in nemogcm_init 
     32 * CALL timing_finalize at the end of nemoggcm 
     33 * at the end of step, IF( kt == nit000) CALL timing_reset (once the list of varibles has been built) 
     34 * in each routine to instrument : CALL timing_start('NAME')    CALL timing_stop('NAME') 
     35Imbricated sub-sections are allowed and their time is then subtracted from the mother section unless the call of timing of the section is done in the following way CALL timing_start('NAME')    CALL timing_stop('NAME',section)  
     36 
     37==== 2- Auto-assignement ==== 
     38==== 3- Dynamic dynamic memory ==== 
    2739 
    2840