Changeset 10335
- Timestamp:
- 2018-11-19T15:25:00+01:00 (5 years ago)
- Location:
- utils/tools/MPP_PREP
- Files:
-
- 2 added
- 2 edited
-
README.rst (added)
-
find_layout.py (added)
-
namelist (modified) (2 diffs)
-
src/mpp_optimiz_zoom_nc.f90 (modified) (1 diff)
Legend:
- Unmodified
- Added
- Removed
-
utils/tools/MPP_PREP/namelist
r2143 r10335 1 1 !>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2 ! MPP_OPTIMIZ _ZOOM namelist3 ! --------------------------- 2 ! MPP_OPTIMIZE namelist template 3 ! ------------------------------- 4 4 !>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 5 5 ! … … 7 7 ! namspace spatial indexes 8 8 !,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 9 ! NAMELIST /namspace/ jpiglo,jpjglo,jpidta,jpjdta,nizoom,njzoom 10 ! jpiglo = overall size of the zoomed region (i-direction) 11 ! jpjglo = overall size of the zoomed region (j-direction) 12 ! jpidta = overall size of the domain (i-direction) 13 ! jpjdta = overall size of the domain (j-direction) 14 ! nizoom = i -index of point (1,1) of the zoomed region/ jpidta 15 ! njzoom = j -index of point (1,1) of the zoomed region/ jpjdta 16 ! 17 &NAMSPACE 18 jpk=31 19 jpiglo = 182 20 jpjglo = 149 21 jpidta = 182 22 jpjdta = 149 23 nizoom = 1 24 njzoom = 1 9 &namspace 10 nn_jpk = 75 ! number of vertical level 11 nn_izoom = 1 ! i-index of point (1,1) of the zoomed region/ jpidta 12 nn_jzoom = 1 ! j-index of point (1,1) of the zoomed region/ jpjdta 25 13 / 26 14 !''''''''''''''''''''''''''''''''''''' 27 15 ! namproc 28 16 !'''''''''''''''''''''''''''''''''''' 29 ! jprocx = maximum number of proc 30 ! jpmem = 0 : dont care about the use of memory 31 ! 1 : try to optimize the use of memory 32 ! 33 &NAMPROC 34 jprocx= 80 35 jpmem = 0 17 &namproc 18 nn_procmax = 4000 ! maximum number of proc to look for 19 nn_procmin = 100 ! minimum number of proc 20 ln_memchk = .false. ! optimization of memory 36 21 / 37 22 !'''''''''''''''''''''''''''''''''''''' 38 23 ! namparam 39 24 !'''''''''''''''''''''''''''''''''''''' 40 ! ppmcal = memoire en octet d'un processeur 41 ! ppmin = ?? 42 ! ppmax = ?? 43 ! 44 &NAMPARAM 45 ppmcal = 225000000. 46 ppmin = 0.4 47 ppmax = 0.9 25 &namparam 26 rn_ppmcal = 225000000. ! maximum memory for 1 processor 27 rn_ppmin = 0.4 ! minimum ratio for filling the available memory 28 rn_ppmax = 0.9 ! maximum ratio for filling the available memory 48 29 / 49 !50 30 !''''''''''''''''''''''''''''''''''''''' 51 31 ! namfile of filename 52 32 !,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 53 ! NAMELIST /namfile/ cbathy 54 ! cbathy = name of the bathymetric file(nc) 55 ! ln_zps = false if bathy file is level 56 ! true if bathy file is meter 57 ! 58 &NAMFILE 59 cbathy='bathymetry_depth_ORCA_R2_V3.nc' 60 ln_zps=.true. 33 &namfile 34 cn_fbathy = 'domain_cfg.nc' ! bathy file name 35 cn_var = 'bottom_level' ! Bathy variable name 36 cn_x = 'x' ! bathy x dimension name 37 cn_y = 'y' ! bathy y dimension name 38 ln_zps = .true. ! partial step flag 61 39 / 62 40 ! 63 41 !'''''''''''''''''''''''''''''''''''''' 64 ! namkeep jpni jpnj kept42 ! namkeep option -keep. Specify the root name of the overdata file 65 43 !,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 66 ! NAMELIST /namkeep/ jpni,jpnj,covdta 67 ! jpni = number of procs. in the i-direction 68 ! jpnj = number of procs. in the j-direction 69 ! covdta = Root for the overdata file name . 70 ! Complete name will be {covdta}.{jpni}x{jpnj}_{jpnij} 71 &NAMKEEP 72 jpni = 7 73 jpnj =11 74 covdta = 'ORCA2-zahir' 44 &namkeep 45 cn_fovdta = 'domain_cfg' ! Root for the overdata file name 46 ! complete name will be {covdta}.{jpni}x{jpnj}_{jpnij} 75 47 / -
utils/tools/MPP_PREP/src/mpp_optimiz_zoom_nc.f90
r6412 r10335 1 PROGRAM mpp_optimiz_nc 2 !!--------------------------------------------------------------------- 3 !! 4 !! PROGRAM MPP_OPTIMIZ_NC 5 !! *********************** 6 !! 7 !! PURPOSE : 8 !! --------- 9 !! This program is build to optimize the domain beakdown into 10 !! subdomain for mpp computing. 11 !! Once the grid size, and the land/sea mask is known, it looks 12 !! for all the possibilities within a range of setting parameters 13 !! and determine the optimal. 14 !! 15 !! Optimization is done with respect to the maximum number of 16 !! sea processors and to the maximum numbers of procs (jprocx) 17 !! 18 !! Optional optimization can be performed takink into account 19 !! the maximum available processor memory ppmcal. This is 20 !! activated if jpmen =1 21 !! 22 !! history: 23 !! -------- 24 !! original : 95-12 (Imbard M) for OPA8.1, CLIPPER 25 !! f90 : 03-06 (Molines JM), namelist as input 26 !! : 05-05 (Molines JM), bathy in ncdf 27 !!---------------------------------------------------------------------- 28 !! * modules used 29 USE netcdf 30 31 IMPLICIT NONE 32 33 INTEGER :: jprocx=250 !: maximum number of proc. (Read from namelist) 34 INTEGER :: jpmem=0 !: memory constraint (1) or no constraint (0) 35 ! ! (use 1 with caution as the memory size of 36 ! ! the code lays on OPA 8.1 estimates ...) 37 ! 38 INTEGER :: & 39 jpk = 46 , & !: vertical levels (namelist) 40 jpiglo = 1442, & !: I-size of the model (namelist) 41 jpjglo = 1021, & !: J-size of the model (namelist) 42 jpidta = 1442, & !: I-size of the data file (namelist) 43 jpjdta = 1021, & !: J-size of the data files (namelist) 44 nizoom = 1 , & !: I zoom indicator (namelist) 45 njzoom = 1 , & !: J zoom indicatori (namelist) 46 numnam = 4 !: logical unit for the namelist 47 NAMELIST /namspace/ jpk,jpiglo,jpjglo,jpidta,jpjdta,nizoom,njzoom 48 NAMELIST /namproc/ jprocx, jpmem 49 50 INTEGER :: jpnix ,jpnjx 1 PROGRAM mpp_optimize 2 !!====================================================================== 3 !! *** PROGRAM mpp_optimize *** 4 !!===================================================================== 5 !! ** Purpose : Propose possible domain decompositions for a given 6 !! bathymetric file, which is particularly intersting when 7 !! we want to eliminate land-only domain. 8 !! All solution are proposed and written to output file. 9 !! The ratio between the effective number of computed 10 !! point and the total number of points in the domain is 11 !! given and is probably a major criteria for choosing a 12 !! domain decomposition. 13 !! 14 !! ** Method : Use mpp_init like code for seting up the decomposition 15 !! and evaluate the efficiency of the decomposition. 16 !! History 17 !! original : 95-12 (Imbard M) for OPA8.1, CLIPPER 18 !! f90 : 03-06 (Molines JM), namelist as input 19 !! : 05-05 (Molines JM), bathy in ncdf 20 !! : 13-03 (Molines JM), Nemo-like coding and license. 21 !! : 18-10 (Mathiot P), upgrade the NEMO 4.0 22 !!---------------------------------------------------------------------- 23 !!---------------------------------------------------------------------- 24 !! routines : description 25 !!---------------------------------------------------------------------- 26 27 28 !!---------------------------------------------------------------------- 29 !! MPP-PREP, MEOM 2013 30 !! $Id$ 31 !! Copyright (c) 2013, J.-M. Molines 32 !! Software governed by the CeCILL licence (Licence/MPP-PREPCeCILL.txt) 33 !!---------------------------------------------------------------------- 34 USE netcdf 35 36 IMPLICIT NONE 37 38 INTEGER, PARAMETER :: jpreci=1 ,jprecj=1 !: overlap between processors 39 40 ! Namelist declaration and definition 41 ! ----------------------------------- 42 INTEGER :: nn_procmax =250 !: maximum number of proc. (Read from namelist) 43 INTEGER :: nn_procmin = 1 !: maximum number of proc. (Read from namelist) 44 LOGICAL :: ln_memchk = .FALSE. ! add a memory constraint if true (obsolete) 45 NAMELIST /namproc/ nn_procmax, nn_procmin, ln_memchk 46 ! 47 INTEGER :: nn_jpk = 46 !: vertical levels 48 INTEGER :: nn_izoom = 1 !: I zoom indicator 49 INTEGER :: nn_jzoom = 1 !: J zoom indicator 50 NAMELIST /namspace/ nn_jpk, nn_izoom, nn_jzoom 51 ! 52 ! Following variables are used only if ln_memchk=.true. 53 REAL(KIND=4) :: required_memory, rppmpt !: not in namelist working array 54 REAL(KIND=4) :: rn_ppmcal = 225000000. !: maximum memory of one processor for a 55 !: given machine (in 8 byte words) 56 REAL(KIND=4) :: rn_ppmin = 0.4 !: minimum ratio to fill the memory 57 REAL(KIND=4) :: rn_ppmax = 0.9 !: maximum ratio to fill the memory 58 NAMELIST /namparam/ rn_ppmcal, rn_ppmin, rn_ppmax 59 ! 60 CHARACTER(LEN=80) :: cn_var='none' !: Variable name of the bathymetry 61 CHARACTER(LEN=80) :: cn_x='x' !: X dimension name 62 CHARACTER(LEN=80) :: cn_y='y' !: Y dimension name 63 CHARACTER(LEN=80) :: cn_fbathy !: File name of the netcdf bathymetry (namelist) 64 LOGICAL :: ln_zps=.FALSE. !: Logical flag for partial cells. 65 NAMELIST /namfile/ cn_fbathy, cn_var, cn_x, cn_y, ln_zps 66 ! 67 CHARACTER(LEN=80) :: cn_fovdta !: root file name for keep output 68 NAMELIST /namkeep/ cn_fovdta 69 ! 70 INTEGER :: numnam = 4 ! logical unit for namelist 71 INTEGER :: numout = 10 ! logical unit for output 72 INTEGER :: npiglo, npjglo ! domain size 73 INTEGER :: npidta, npjdta ! domain size 74 75 INTEGER :: ji, jj, jni, jnj ! dummy loop index 76 INTEGER :: ii, ij, jjc ! dummy loop index 77 INTEGER :: narg, iargc, ijarg ! browsing command line 78 79 ! Decomposition related arrays (using same meaning than in NEMO) 80 INTEGER, DIMENSION(:,:), ALLOCATABLE :: ilci, ilcj ,iimppt, ijmppt 81 INTEGER, DIMENSION(:) , ALLOCATABLE :: nlei_ocea, nldi_ocea 82 INTEGER, DIMENSION(:) , ALLOCATABLE :: nlej_ocea, nldj_ocea 83 INTEGER, DIMENSION(:) , ALLOCATABLE :: nlei_land, nldi_land 84 INTEGER, DIMENSION(:) , ALLOCATABLE :: nlej_land, nldj_land 85 INTEGER :: nimpp, njmpp 86 INTEGER :: nreci, nrecj 87 INTEGER :: ili, ilj 88 INTEGER :: jarea, iarea, iarea0 89 INTEGER :: iresti, irestj 90 ! 91 INTEGER :: ioce, isurf !: number of ocean points cumulated, per_proc 92 INTEGER :: ioce_opt !: number of ocean points cumulated for optimal case 93 INTEGER :: nland, nocea, nvalid !: number of land, ocean, memory_valid procs 94 INTEGER :: nland_opt !: optimal number of land procs 95 INTEGER :: ii1, ii2, ij1, ij2 !: limit of subdomain in global domain 96 INTEGER :: jpimax, jpjmax !: size of sub domain 97 INTEGER :: jpimax_opt, jpjmax_opt !: size of sub domain for optimal case 98 INTEGER :: inf10, inf30, inf50 !: 99 INTEGER :: inf10_opt, inf30_opt, inf50_opt !: in optimal case 100 INTEGER :: npni_opt, npnj_opt !: optimal domain decomposition 101 102 INTEGER :: iminproci, imaxproci !: limits of the processor loop 103 INTEGER :: iminprocj, imaxprocj !: can be reduded to nkeepi, nkeepj 104 105 ! Saving criteria 106 REAL(KIND=4) :: ratio_min=99999. !: keep only decomposition with ration less than ratio_min 107 INTEGER :: nocea_min = 1 !: minimum number of ocean procs for saving 108 INTEGER :: nmodulo = 1 !: Only keep solution multiple of nmodulo 109 LOGICAL :: ll_criteria=.TRUE. !: 110 ! 111 REAL(KIND=4) :: oce_cover 112 REAL(KIND=4) :: oce_cover_min, oce_cover_max, ratio 113 REAL(KIND=4) :: oce_cover_min_opt, oce_cover_max_opt, ratio_opt 114 REAL(KIND=4), DIMENSION(:,:), ALLOCATABLE :: tmask ! npiglo x npjglo 115 REAL(KIND=4), DIMENSION(:,:), ALLOCATABLE :: bathy ! npidta x npjdta 116 117 ! CDF stuff 118 INTEGER :: ncid, istatus, id 119 LOGICAL :: ll_good = .FALSE. 120 121 CHARACTER(LEN=80) :: cf_namlist='namelist' 122 CHARACTER(LEN=80) :: cf_out='processor.layout' 123 CHARACTER(LEN=80) :: cdum ! dummy character variable 124 125 ! Keep stuff 126 LOGICAL :: ll_keep = .FALSE. 127 INTEGER :: nkeepi, nkeepj !: for option -keep : the retained decomposition 51 128 ! 52 INTEGER,PARAMETER :: jpreci=1 ,jprecj=1 53 ! 54 ! Following variables are used only if jpmem=1 55 REAL(KIND=4) :: ppmpt , & 56 ppmcal = 225000000., & !: maximum memory of one processor for a given machine (in 8 byte words) 57 ppmin = 0.4, & !: minimum ratio to fill the memory 58 ppmax = 0.9 !: maximum ration to fill the memory 59 ! Aleph 60 ! PARAMETER(ppmcal= 16000000.) 61 !Brodie 62 ! PARAMETER(ppmcal=250000000.) 63 ! Uqbar 64 ! PARAMETER(ppmcal=3750000000.) 65 ! Zahir 66 ! PARAMETER(ppmcal=225000000.) 67 68 CHARACTER(LEN=80) :: cbathy, & !: File name of the netcdf bathymetry (namelist) 69 & clvar !: Variable name in netcdf for the bathy to be read 70 LOGICAL :: ln_zps=.false. !: Logical flag for partial cells. 71 NAMELIST /namfile/ cbathy, ln_zps 72 NAMELIST /namparam/ ppmcal, ppmin, ppmax 73 ! 74 INTEGER :: iumout = 1 75 INTEGER :: ji,jj,jn,jni,jnj,jni2,jnj2 76 INTEGER :: iumbat,ifreq,il1,il2 77 INTEGER :: ii,iim,ij,ijm,imoy,iost,iresti,irestj,isurf,ivide 78 INTEGER :: iilb,ijlb,ireci,irecj,in 79 INTEGER :: ipi,ipj 80 INTEGER :: inf10,inf30,inf50,iptx,isw 81 INTEGER :: iii,iij,iiii,iijj,iimoy,iinf10,iinf30,iinf50 82 ! 83 INTEGER,DIMENSION(:,:),ALLOCATABLE :: ibathy ! jpidta -jpjdta 84 INTEGER,DIMENSION(:,:),ALLOCATABLE :: ippdi, ippdj ,iidom, ijdom 85 ! 86 REAL(KIND=4) :: zmin,zmax,zper,zmem 87 REAL(KIND=4) :: zzmin,zzmax,zperx 88 REAL(KIND=4),DIMENSION(:,:),ALLOCATABLE :: zmask ,& ! jpiglo -jpjglo 89 & zdta ! jpidta -jpjdta 90 91 ! CDF stuff 92 INTEGER :: ncid, ivarid, istatus 93 LOGICAL :: llbon=.false. 94 ! 95 ! 0. Initialisation 96 ! ----------------- 97 OPEN(numnam,FILE='namelist') 98 REWIND(numnam) 99 READ(numnam,namspace) 100 101 REWIND(numnam) 102 READ(numnam,namfile) 103 104 REWIND(numnam) 105 READ(numnam,namparam) 106 107 REWIND(numnam) 108 READ(numnam,namproc) 109 110 ! estimated code size expressed in number of 3D arrays (valid for OPA8.1) 111 ppmpt = 55.+73./jpk 112 jpnix = jprocx ; jpnjx=jprocx 113 114 ALLOCATE ( ibathy(jpidta,jpjdta), zmask(jpiglo,jpjglo),zdta(jpidta,jpjdta) ) 115 ALLOCATE (ippdi(jpnix,jpnjx), ippdj(jpnix,jpnjx) ) 116 ALLOCATE (iidom(jpnix,jpnjx), ijdom(jpnix,jpnjx) ) 117 118 OPEN(iumout,FILE='processor.layout') 119 WRITE(iumout,*) 120 WRITE(iumout,*) ' optimisation de la partition' 121 WRITE(iumout,*) ' ----------------------------' 122 WRITE(iumout,*) 123 ! 124 ! * Read cdf bathy file 125 ! 126 IF ( ln_zps ) THEN ! partial steps 127 clvar = 'Bathymetry' 128 ELSE 129 clvar = 'Bathy_level' ! full steps 129 130 !!---------------------------------------------------------------------- 131 narg=iargc() 132 ijarg=1 133 IF ( narg == 0 ) THEN 134 PRINT *,' try mpp_optimize -h for instructions !' 135 STOP 136 ENDIF 137 ! 138 DO WHILE ( ijarg <= narg ) 139 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 140 SELECT CASE ( cdum ) 141 CASE ('-h') 142 PRINT *,' usage : mpp_optimize [ -h ] [-keep jpni jpnj] [ -o file out ] ' 143 PRINT *,' [ -modulo val ] [-r ratio] [-minocean procs] -n namelist' 144 PRINT *,' ' 145 PRINT *,' PURPOSE :' 146 PRINT *,' This program is build to optimize the domain beakdown into' 147 PRINT *,' subdomain for mpp computing.' 148 PRINT *,' Once the grid size, and the land/sea mask is known, it looks' 149 PRINT *,' for all the possibilities within a range of setting parameters' 150 PRINT *,' and determine the optimal.' 151 PRINT *,'' 152 PRINT *,' Optimization is done with respect to the maximum number of' 153 PRINT *,' sea processors and to the maximum numbers of procs (nn_procmax)' 154 PRINT *,' ' 155 PRINT *,' Optional optimization can be performed taking into account' 156 PRINT *,' the maximum available processor memory rn_ppmcal. This is' 157 PRINT *,' activated if ln_memchk is set true in the namelist' 158 PRINT *,' ' 159 PRINT *,' Additional criteria can be given on the command line to reduce' 160 PRINT *,' the amount of possible choices.' 161 PRINT *,' ' 162 PRINT *,' ARGUMENTS :' 163 PRINT *,' -n namelist : indicate the name of the namelist to use' 164 PRINT *,' ' 165 PRINT *,' OPTIONS :' 166 PRINT *,' -h : print this help message' 167 PRINT *,' -keep jpni jpnj : print a file suitable for plotting,' 168 PRINT *,' corresponding to the given decomposition' 169 PRINT *,' -o output file : give the name of the output file' 170 PRINT *,' default is ',TRIM(cf_out) 171 PRINT *,' -modulo val : only retain decomposition whose total number' 172 PRINT *,' of util processors (sea) are a multiple of val' 173 PRINT *,' -r ratio : only retain decomposition with a ratio computed/global' 174 PRINT *,' less or equal to the given ratio' 175 PRINT *,' -minocean procs : only retain decomposition with a number of ' 176 PRINT *,' ocean procs greater of equal to procs' 177 PRINT *,' ' 178 PRINT *,' REQUIRED FILES :' 179 PRINT *,' A bathymetric file and an ad-hoc namelist are required.' 180 PRINT *,' The file name of the bathymetry is specified in the namelist' 181 PRINT *,' ' 182 PRINT *,' OUTPUT : ' 183 PRINT *,' ',TRIM(cf_out),' : an ascii file with all found possibilities' 184 PRINT *,' ' 185 STOP 186 CASE ('-n' ) 187 CALL getarg(ijarg,cf_namlist) ; ijarg=ijarg+1 188 CASE ('-o' ) 189 CALL getarg(ijarg,cf_out) ; ijarg=ijarg+1 190 CASE ('-keep' ) 191 ll_keep=.TRUE. 192 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 ; READ( cdum,*) nkeepi 193 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 ; READ( cdum,*) nkeepj 194 CASE ('-modulo' ) 195 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 ; READ( cdum,*) nmodulo 196 CASE ('-r' ) 197 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 ; READ( cdum,*) ratio_min 198 CASE ('-minocean' ) 199 CALL getarg(ijarg,cdum) ; ijarg=ijarg+1 ; READ( cdum,*) nocea_min 200 END SELECT 201 ENDDO 202 203 ! Open and read the namelist 204 OPEN(numnam,FILE=cf_namlist) 205 REWIND(numnam) 206 READ(numnam,namspace) 207 208 REWIND(numnam) 209 READ(numnam,namfile) 210 211 REWIND(numnam) 212 READ(numnam,namparam) 213 214 REWIND(numnam) 215 READ(numnam,namproc) 216 217 REWIND(numnam) 218 READ(numnam,namkeep) ! only used for -keep option but still ... 219 CLOSE(numnam) 220 221 ! estimated code size expressed in number of 3D arrays (valid for OPA8.1) to be tuned for OPA9.0/Nemo 222 rppmpt = 55.+73./nn_jpk 223 224 ! Open bathy file an allocate required memory 225 INQUIRE( FILE=cn_fbathy, EXIST=ll_good ) 226 IF( ll_good ) THEN 227 istatus = NF90_OPEN(cn_fbathy, NF90_NOWRITE, ncid) 228 istatus = NF90_INQ_DIMID(ncid, cn_x, id) ; istatus = NF90_INQUIRE_DIMENSION(ncid, id, len=npiglo) 229 istatus = NF90_INQ_DIMID(ncid, cn_y, id) ; istatus = NF90_INQUIRE_DIMENSION(ncid, id, len=npjglo) 230 npidta = npiglo ; npjdta=npjglo 231 ELSE 232 PRINT *,' File missing : ', TRIM(cn_fbathy) 233 STOP 42 234 ENDIF 235 236 ALLOCATE (tmask(npiglo,npjglo), bathy(npidta,npjdta) ) 237 ALLOCATE (ilci(nn_procmax,nn_procmax), ilcj(nn_procmax,nn_procmax) ) 238 ALLOCATE (iimppt(nn_procmax,nn_procmax), ijmppt(nn_procmax,nn_procmax) ) 239 240 ! Open output file for results 241 IF ( ll_keep ) THEN 242 nn_procmax = nkeepi*nkeepj ! reduce nn_procmax 243 ! File will be open later 244 ELSE 245 OPEN(numout,FILE=cf_out) 246 WRITE(numout,*) 247 WRITE(numout,*) ' Domain decomposition optimization ' 248 WRITE(numout,*) ' ----------------------------------' 249 WRITE(numout,*) 250 ENDIF 251 ! 252 ALLOCATE ( nlei_ocea(nn_procmax), nldi_ocea(nn_procmax), nlej_ocea(nn_procmax), nldj_ocea(nn_procmax) ) 253 ALLOCATE ( nlei_land(nn_procmax), nldi_land(nn_procmax), nlej_land(nn_procmax), nldj_land(nn_procmax) ) 254 ! 255 ! Read cdf bathy file 256 IF ( cn_var == 'none' ) THEN ! automatic detection of variable name according to partial step 257 IF ( ln_zps ) THEN ! partial steps 258 cn_var = 'Bathymetry' 259 ELSE 260 cn_var = 'Bathy_level' ! full steps 261 ENDIF 262 ENDIF 263 PRINT *,'' 264 PRINT *,' ocean/land file used is: ', TRIM(cn_fbathy) 265 PRINT *,' variable used to find ocean domain is: ', TRIM(cn_var) 266 PRINT *,' Dimensions (jpi x jpj) are: ',npiglo,'x',npjglo 267 PRINT *,'' 268 269 istatus = NF90_INQ_VARID (ncid, cn_var, id) 270 istatus = NF90_GET_VAR (ncid, id, bathy) 271 istatus = NF90_CLOSE (ncid) 272 ! 273 ! Building the mask ( eventually on a smaller domain than the bathy) 274 tmask(:,:) = bathy(nn_izoom:nn_izoom+npiglo -1, nn_jzoom:nn_jzoom+npjglo -1) 275 276 WHERE ( tmask > 0 ) 277 tmask = 1. 278 ELSEWHERE 279 tmask = 0. 280 ENDWHERE 281 282 ! Main loop on processors 283 ! ------------------------ 284 ! initialization of working variables 285 npni_opt=1 ; npnj_opt=1 286 jpimax_opt=npiglo ; jpjmax_opt=npjglo 287 nland_opt=0 288 ioce_opt=0 289 oce_cover_min_opt=0. ; oce_cover_max_opt=0. 290 inf10_opt=0 ; inf30_opt=0 ; inf50_opt=0 291 ratio_opt=1. 292 293 nvalid=0 ! counter for valid case ( ln_memchk true ) 294 IF ( ll_keep ) THEN 295 iminproci = nkeepi ; imaxproci = iminproci 296 iminprocj = nkeepj ; imaxprocj = iminprocj 297 ELSE 298 iminproci = 1 ; imaxproci = MIN( nn_procmax, npiglo ) 299 iminprocj = 1 ; imaxprocj = MIN( nn_procmax, npjglo ) 300 ENDIF 301 302 ! loop on all decomposition a priori 303 PRINT *, 'Loop over all the decomposition (can take a while) ...' 304 PRINT *, '' 305 DO jni=iminproci, imaxproci 306 DO jnj=iminprocj, imaxprocj 307 ! Limitation of the maxumun number of PE's 308 IF ( jni*jnj <= nn_procmax .AND. jni*jnj >= nn_procmin ) THEN 309 ! 310 ! 1. Dimension arrays for subdomains 311 ! ----------------------------------- 312 ! 313 ! Partition : size of sub-domain 314 jpimax=(npiglo-2*jpreci + (jni-1))/jni + 2*jpreci 315 jpjmax=(npjglo-2*jprecj + (jnj-1))/jnj + 2*jprecj 316 ! 317 ! Memory optimization ? 318 IF ( ln_memchk ) THEN 319 required_memory=rppmpt*jpimax*jpjmax*nn_jpk 320 IF( required_memory > rn_ppmcal ) EXIT 321 IF( required_memory > rn_ppmax*rn_ppmcal .OR. required_memory < rn_ppmin*rn_ppmcal) EXIT 322 ENDIF 323 nvalid=nvalid+1 324 ! 325 ! Position of each sub domain (jni x jni in total ) 326 nreci = 2*jpreci ; nrecj = 2*jprecj 327 iresti = 1 + MOD ( npiglo - nreci - 1 , jni ) ; irestj = 1 + MOD ( npjglo - nrecj - 1 , jnj ) 328 ! 329 ! 330 ilci( 1:iresti, 1:jnj) = jpimax 331 ilci(iresti+1:jni , 1:jnj) = jpimax-1 332 333 ilcj(1:jni, 1:irestj) = jpjmax 334 ilcj(1:jni,irestj+1:jnj ) = jpjmax-1 335 336 ! 2. Index arrays for subdomains 337 ! ------------------------------- 338 iimppt(1:jni, 1:jnj) = 1 339 ijmppt(1:jni, 1:jnj) = 1 340 IF( jni > 1 ) THEN 341 DO jj=1,jnj 342 DO ji=2,jni 343 iimppt(ji,jj)= iimppt(ji-1,jj) + ilci(ji-1,jj) - nreci 344 END DO 345 END DO 346 ENDIF 347 348 IF( jnj > 1 ) THEN 349 DO jj=2,jnj 350 DO ji=1,jni 351 ijmppt(ji,jj)= ijmppt(ji,jj-1) + ilcj(ji,jj-1) - nrecj 352 END DO 353 END DO 354 ENDIF 355 ! 356 ! Loop on each subdomain to look for land proportion 357 nland = 0 358 nocea = 0 359 ioce = 0 360 oce_cover_min = 1.e+20 361 oce_cover_max = -1.e+20 362 inf10=0 363 inf30=0 364 inf50=0 365 ! 366 ! 3. Subdomain description in the Regular Case 367 ! -------------------------------------------- 368 ! 369 DO jarea = 1, jni*jnj 370 iarea0 = jarea - 1 371 ii = 1 + MOD(iarea0,jni) 372 ij = 1 + iarea0/jni 373 ili = ilci(ii,ij) 374 ilj = ilcj(ii,ij) 375 376 isurf = 0 377 ! loop on inner point of sub-domain 378 DO jj=1, ilj 379 DO ji=1, ili 380 IF( tmask(ji + iimppt(ii,ij) - 1, jj + ijmppt(ii,ij) - 1) == 1 ) isurf=isurf+1 381 END DO 382 END DO 383 384 nimpp = iimppt(ii,ij) 385 njmpp = ijmppt(ii,ij) 386 ii1 = nimpp+jpreci ; ii2 = nimpp+ili-1 -jpreci 387 ij1 = njmpp+jprecj ; ij2 = njmpp+ilj-1 -jprecj 388 IF ( isurf == 0 ) THEN 389 nland = nland+1 390 nldi_land(nland) = ii1 391 nlei_land(nland) = ii2 392 nldj_land(nland) = ij1 393 nlej_land(nland) = ij2 394 ELSE 395 nocea = nocea+1 396 ioce = ioce + isurf 397 nldi_ocea(nocea) = ii1 398 nlei_ocea(nocea) = ii2 399 nldj_ocea(nocea) = ij1 400 nlej_ocea(nocea) = ij2 401 ENDIF 402 403 ! ratio of wet points over total number of point per proc. 404 oce_cover = float(isurf)/float(jpimax*jpjmax) 405 406 IF(oce_cover_min > oce_cover .AND. isurf /= 0) oce_cover_min=oce_cover 407 IF(oce_cover_max < oce_cover .AND. isurf /= 0) oce_cover_max=oce_cover 408 IF(oce_cover < 0.1 .AND. isurf /= 0) inf10=inf10+1 409 IF(oce_cover < 0.3 .AND. isurf /= 0) inf30=inf30+1 410 IF(oce_cover < 0.5 .AND. isurf /= 0) inf50=inf50+1 411 ! 412 !END DO ! loop on processors 413 END DO ! loop on processors 414 ! 415 ratio=float(nocea)*float(jpimax*jpjmax)/float(npiglo*npjglo) 416 417 ! criteria for printing results 418 ll_criteria = ( ( MOD ( nocea, nmodulo ) == 0 ) .AND. & 419 & ( ratio <= ratio_min ) .AND. & 420 & ( nocea >= nocea_min ) ) 421 IF ( ll_keep ) THEN ! the loop in done only once ! 422 WRITE(cdum,'(a,"-",i3.3,"x",i3.3,"_",i4.4)') TRIM(cn_fovdta), nkeepi, nkeepj, nocea 423 OPEN(numout, file=cdum ) 424 WRITE(numout,'("# ocean ",i5)') nocea 425 DO jjc=1, nocea 426 WRITE(numout,'("#",i5)') jjc 427 WRITE(numout,'(2i5)') nldi_ocea(jjc)-1+nn_izoom-1, nldj_ocea(jjc)-1+nn_jzoom -1 428 WRITE(numout,'(2i5)') nlei_ocea(jjc)+1+nn_izoom-1, nldj_ocea(jjc)-1+nn_jzoom -1 429 WRITE(numout,'(2i5)') nlei_ocea(jjc)+1+nn_izoom-1, nlej_ocea(jjc)+1+nn_jzoom -1 430 WRITE(numout,'(2i5)') nldi_ocea(jjc)-1+nn_izoom-1, nlej_ocea(jjc)+1+nn_jzoom -1 431 WRITE(numout,'(2i5)') nldi_ocea(jjc)-1+nn_izoom-1, nldj_ocea(jjc)-1+nn_jzoom -1 432 WRITE(numout,'(2i5)') 9999, 9999 433 ENDDO 434 ! 435 WRITE(numout,'("# land ",i5)') nland 436 DO jjc=1, nland 437 WRITE(numout,'("# land ",i5)') jjc 438 WRITE(numout,'(2i5)') nldi_land(jjc)-1+nn_izoom-1, nldj_land(jjc)-1+nn_jzoom -1 439 WRITE(numout,'(2i5)') nlei_land(jjc)+1+nn_izoom-1, nldj_land(jjc)-1+nn_jzoom -1 440 WRITE(numout,'(2i5)') nlei_land(jjc)+1+nn_izoom-1, nlej_land(jjc)+1+nn_jzoom -1 441 WRITE(numout,'(2i5)') nldi_land(jjc)-1+nn_izoom-1, nlej_land(jjc)+1+nn_jzoom -1 442 WRITE(numout,'(2i5)') nldi_land(jjc)-1+nn_izoom-1, nldj_land(jjc)-1+nn_jzoom -1 443 WRITE(numout,'(2i5)') nlei_land(jjc)+1+nn_izoom-1, nlej_land(jjc)+1+nn_jzoom -1 444 WRITE(numout,'(2i5)') nldi_land(jjc)-1+nn_izoom-1, nlej_land(jjc)+1+nn_jzoom -1 445 WRITE(numout,'(2i5)') nlei_land(jjc)+1+nn_izoom-1, nldj_land(jjc)-1+nn_jzoom -1 446 WRITE(numout,'(2i5)') 9999, 9999 447 ENDDO 448 ! 449 ELSE 450 IF ( ll_criteria ) THEN 451 WRITE(numout,*) ' iresti=',iresti,' irestj=',irestj 452 WRITE(numout,*) '--> Total number of domains ',jni*jnj 453 WRITE(numout,*) ' ' 454 WRITE(numout,*) ' jpni=',jni ,' jpnj=',jnj 455 WRITE(numout,*) ' jpi= ',jpimax ,' jpj= ',jpjmax 456 WRITE(numout,*) ' Number of ocean processors ', nocea 457 WRITE(numout,*) ' Number of land processors ', nland 458 WRITE(numout,*) ' Mean ocean coverage per domain ', float(ioce)/float(nocea)/float(jpimax*jpjmax) 459 WRITE(numout,*) ' Minimum ocean coverage ', oce_cover_min 460 WRITE(numout,*) ' Maximum ocean coverage ', oce_cover_max 461 WRITE(numout,*) ' nb of proc with coverage < 10 % ', inf10 462 WRITE(numout,*) ' nb of proc with coverage 10 < nb < 30 % ', inf30 - inf10 463 WRITE(numout,*) ' nb of proc with coverage 30 < nb < 50 % ', inf50 - inf30 464 WRITE(numout,*) ' Number of computed points ', nocea*jpimax*jpjmax 465 WRITE(numout,*) ' Overhead of computed points ', nocea*jpimax*jpjmax-npiglo*npjglo 466 WRITE(numout,*) ' % sup (computed / global) ', ratio 467 WRITE(numout,*) 468 ENDIF ! note that indication of optimum does not take modulo into account (for information) 469 ! 470 ! Look for optimum 471 IF( nland > nland_opt ) THEN 472 npni_opt = jni 473 npnj_opt = jnj 474 jpimax_opt = jpimax 475 jpjmax_opt = jpjmax 476 nland_opt = nland 477 ioce_opt = ioce 478 oce_cover_min_opt = oce_cover_min 479 oce_cover_max_opt = oce_cover_max 480 inf10_opt = inf10 481 inf30_opt = inf30 482 inf50_opt = inf50 483 ratio_opt = ratio 484 ELSE IF( nland == nland_opt .AND. ratio_opt < ratio) THEN 485 npni_opt = jni 486 npnj_opt = jnj 487 jpimax_opt = jpimax 488 jpjmax_opt = jpjmax 489 ioce_opt = ioce 490 oce_cover_min_opt = oce_cover_min 491 oce_cover_max_opt = oce_cover_max 492 inf10_opt = inf10 493 inf30_opt = inf30 494 inf50_opt = inf50 495 ratio_opt = ratio 496 ENDIF 497 ENDIF 130 498 ENDIF 131 132 INQUIRE( FILE=cbathy, EXIST=llbon )133 IF( llbon ) THEN134 istatus=NF90_OPEN(cbathy,NF90_NOWRITE,ncid)135 istatus=NF90_INQ_VARID(ncid,clvar,ivarid)136 istatus=NF90_GET_VAR(ncid,ivarid,zdta)137 istatus=NF90_CLOSE(ncid)138 ELSE139 PRINT *,' File missing : ', trim(cbathy)140 STOP499 END DO 500 END DO 501 ! 502 ! print optimal result 503 IF ( .NOT. ll_keep ) THEN 504 IF ( nvalid == 0 ) THEN 505 WRITE(numout,*) ' no possible choice ...' 506 WRITE(numout,*) 507 WRITE(numout,*) 'insufficient number of processors for the available memory' 508 STOP 141 509 ENDIF 142 ibathy(:,:)=zdta(:,:) 143 144 ! 145 ! Building the mask 146 DO jj=1,jpjglo 147 DO ji=1,jpiglo 148 zmask(ji,jj) = float(ibathy(ji+nizoom - 1,jj+njzoom -1)) 149 END DO 150 END DO 151 152 DO jj=1,jpjglo 153 DO ji=1,jpiglo 154 zmask(ji,jj)= min(REAL(1.,kind=4),max(REAL(0.,kind=4),zmask(ji,jj))) ! Old vector coding rule ... 155 END DO 156 END DO 157 ! 158 ! Main loop on processors 159 ! ------------------------ 160 iii=1 ; iij=1 161 iiii=jpiglo ; iijj=jpjglo 162 iptx=0 163 iimoy=0 164 zzmin=0. ; zzmax=0. 165 iinf10=0 ; iinf30=0 ; iinf50=0 166 zperx=1. 167 in=0 168 DO jni=1,jpnix 169 DO jnj=1,jpnjx 170 ! 171 ! Limitation ob the maxumun number of PE's 172 IF(jni*jnj > jprocx) goto 1000 173 ! 174 ! Partition 175 ipi=(jpiglo-2*jpreci + (jni-1))/jni + 2*jpreci 176 ipj=(jpjglo-2*jprecj + (jnj-1))/jnj + 2*jprecj 177 ! 178 ! Memory optimization ? 179 isw=0 180 zmem=ppmpt*ipi*ipj*jpk 181 IF(zmem > ppmcal) go to 1000 182 IF(jpmem == 1) THEN 183 IF(zmem.GT.ppmax*ppmcal.OR.zmem.LT.ppmin*ppmcal) isw=1 184 ENDIF 185 IF(isw.EQ.1) go to 1000 186 in=in+1 187 ! 188 WRITE(iumout,*) '--> nombre de processeurs ',jni*jnj 189 WRITE(iumout,*) ' ' 190 WRITE(iumout,*) " jpni=",jni ," jpnj=",jnj 191 WRITE(iumout,*) " jpi= ",ipi ," jpj= ",ipj 192 zper=(jni*jnj*ipi*ipj)/float(jpiglo*jpjglo) 193 WRITE(iumout,*) " rapport jpnij*domain/global domain ",zper 194 ! 195 ! Coin en bas a gauche de chaque processeur 196 ! 197 iilb=1 198 ijlb=1 199 ireci=2*jpreci 200 irecj=2*jprecj 201 iresti = MOD ( jpiglo - ireci , jni ) 202 irestj = MOD ( jpjglo - irecj , jnj ) 203 ! 204 IF (iresti.EQ.0) iresti = jni 205 DO jj=1,jnj 206 DO ji=1,iresti 207 ippdi(ji,jj) = ipi 208 END DO 209 DO ji=iresti+1,jni 210 ippdi(ji,jj) = ipi -1 211 END DO 212 END DO 213 IF (irestj.EQ.0) irestj = jnj 214 DO ji=1,jni 215 DO jj=1,irestj 216 ippdj(ji,jj) = ipj 217 END DO 218 DO jj=irestj+1,jnj 219 ippdj(ji,jj) = ipj -1 220 END DO 221 END DO 222 DO jj=1,jnj 223 DO ji=1,jni 224 iidom(ji,jj)=iilb 225 ijdom(ji,jj)=ijlb 226 END DO 227 END DO 228 WRITE(iumout,*) " iresti=",iresti," irestj=",irestj 229 ! 230 ! 2. Boucle sur les processeurs 231 ! ------------------------------ 232 ! 233 ivide=0 234 imoy=0 235 zmin=1.e+20 236 zmax=-1.e+20 237 inf10=0 238 inf30=0 239 inf50=0 240 ! 241 DO jni2=1,jni 242 DO jnj2=1,jnj 243 244 IF(jni.GT.1)THEN 245 DO jj=1,jnj 246 DO ji=2,jni 247 iidom(ji,jj)=iidom(ji-1,jj)+ippdi(ji-1,jj)-ireci 248 END DO 249 END DO 250 iilb=iidom(jni2,jnj2) 251 ENDIF 252 IF(jnj.GT.1)THEN 253 DO jj=2,jnj 254 DO ji=1,jni 255 ijdom(ji,jj)=ijdom(ji,jj-1)+ippdj(ji,jj-1)-irecj 256 END DO 257 END DO 258 ijlb=ijdom(jni2,jnj2) 259 ENDIF 260 261 ! Check wet points over the entire domain to preserve the MPI communication stencil 262 isurf=0 263 DO jj=1,ippdj(jni2,jnj2) 264 DO ji=1,ippdi(jni2,jnj2) 265 IF(zmask(ji+iilb-1,jj+ijlb-1).EQ.1.) isurf=isurf+1 266 END DO 267 END DO 268 269 IF(isurf.EQ.0) THEN 270 ivide=ivide+1 271 ELSE 272 imoy=imoy+isurf 273 ENDIF 274 zper=float(isurf)/float(ipi*ipj) 275 IF(zmin.GT.zper.AND.isurf.NE.0) zmin=zper 276 IF(zmax.LT.zper.AND.isurf.NE.0) zmax=zper 277 IF(zper.LT.0.1.AND.isurf.NE.0) inf10=inf10+1 278 IF(zper.LT.0.3.AND.isurf.NE.0) inf30=inf30+1 279 IF(zper.LT.0.5.AND.isurf.NE.0) inf50=inf50+1 280 ! 281 ! 282 ! 3. Fin de boucle sur les processeurs, impression 283 ! ------------------------------------------------ 284 ! 285 END DO 286 END DO 287 WRITE(iumout,*) ' nombre de processeurs ',jni*jnj 288 WRITE(iumout,*) ' nombre de processeurs mer ',jni*jnj-ivide 289 WRITE(iumout,*) ' nombre de processeurs terre ',ivide 290 WRITE(iumout,*) ' moyenne de recouvrement ',float(imoy)/float(jni*jnj-ivide)/float(ipi*ipj) 291 WRITE(iumout,*) ' minimum de recouvrement ',zmin 292 WRITE(iumout,*) ' maximum de recouvrement ',zmax 293 WRITE(iumout,*) ' nb de p recouvrement < 10 % ',inf10 294 WRITE(iumout,*) ' nb de p 10 < nb < 30 % ',inf30-inf10 295 WRITE(iumout,*) ' nb de p 30 < nb < 50 % ',inf50-inf10 -inf30 296 WRITE(iumout,*) ' nombre de points integres ', (jni*jnj-ivide)*ipi*ipj 297 WRITE(iumout,*) ' nbr de pts supplementaires ', (jni*jnj-ivide)*ipi*ipj-jpiglo*jpjglo 298 zper=float((jni*jnj-ivide))*float(ipi*ipj)/float(jpiglo*jpjglo) 299 WRITE(iumout,*) ' % sup ',zper 300 WRITE(iumout,*) 301 ! 302 ! 303 ! 4. Recherche de l optimum 304 ! ------------------------- 305 ! 306 IF(ivide.GT.iptx) THEN 307 iii=jni 308 iij=jnj 309 iiii=ipi 310 iijj=ipj 311 iptx=ivide 312 iimoy=imoy 313 zzmin=zmin 314 zzmax=zmax 315 iinf10=inf10 316 iinf30=inf30 317 iinf50=inf50 318 zperx=zper 319 ELSE IF(ivide.EQ.iptx.AND.zperx.LT.zper) THEN 320 iii=jni 321 iij=jnj 322 iiii=ipi 323 iijj=ipj 324 iimoy=imoy 325 zzmin=zmin 326 zzmax=zmax 327 iinf10=inf10 328 iinf30=inf30 329 iinf50=inf50 330 zperx=zper 331 ENDIF 332 ! 333 ! 5. Fin de boucle sur le nombre de processeurs 334 ! --------------------------------------------- 335 ! 336 1000 continue 337 END DO 338 END DO 339 ! 340 ! 341 ! 6. Affichage resultat 342 ! --------------------- 343 ! 344 IF(in.EQ.0) THEN 345 WRITE(iumout,*) ' le choix n'' a pas pu etre fait ' 346 WRITE(iumout,*) 347 WRITE(iumout,*) 'le nombre de processeurs maximum est insuffisant' 348 STOP 349 ENDIF 350 WRITE(iumout,*) ' choix optimum' 351 WRITE(iumout,*) ' =============' 352 WRITE(iumout,*) 353 WRITE(iumout,*) '--> nombre de processeurs ',iii*iij 354 WRITE(iumout,*) ' ' 355 WRITE(iumout,*) " jpni=",iii ," jpnj=",iij 356 WRITE(iumout,*) " jpi= ",iiii ," jpj= ",iijj 357 WRITE(iumout,*) 358 WRITE(iumout,*) ' nombre de processeurs mer ',iii*iij-iptx 359 WRITE(iumout,*) ' nombre de processeurs terre ',iptx 360 WRITE(iumout,*) ' moyenne de recouvrement ',float(iimoy)/float(iii*iij-iptx)/float(iiii*iijj) 361 WRITE(iumout,*) ' minimum de recouvrement ',zzmin 362 WRITE(iumout,*) ' maximum de recouvrement ',zzmax 363 WRITE(iumout,*) ' nb de p recouvrement < 10 % ',iinf10 364 WRITE(iumout,*) ' nb de p 10 < nb < 30 % ',iinf30-iinf10 365 WRITE(iumout,*) ' nb de p 30 < nb < 50 % ',iinf50-iinf10 -iinf30 366 WRITE(iumout,*) ' nombre de points integres ', (iii*iij-iptx)*iiii*iijj 367 WRITE(iumout,*) ' nbr de pts supplementaires ', (iii*iij-iptx)*iiii*iijj-jpiglo*jpjglo 368 WRITE(iumout,*) ' % sup ',zperx 369 WRITE(iumout,*) 370 CLOSE(iumout) 371 ! 372 ! 373 ! 374 STOP 375 END PROGRAM mpp_optimiz_nc 510 511 WRITE(numout,*) ' Optimal choice' 512 WRITE(numout,*) ' ==============' 513 WRITE(numout,*) 514 WRITE(numout,*) '--> Total number of domains ',npni_opt*npnj_opt 515 WRITE(numout,*) ' ' 516 WRITE(numout,*) ' jpni=',npni_opt ,' jpnj=',npnj_opt 517 WRITE(numout,*) ' jpi= ',jpimax_opt ,' jpj= ',jpjmax_opt 518 WRITE(numout,*) 519 WRITE(numout,*) ' Number of ocean processors ', npni_opt*npnj_opt-nland_opt 520 WRITE(numout,*) ' Number of land processors ', nland_opt 521 WRITE(numout,*) ' Mean ocean coverage ', float(ioce_opt)/float(npni_opt*npnj_opt-nland_opt)/float(jpimax_opt*jpjmax_opt) 522 WRITE(numout,*) ' Minimum ocean coverage ', oce_cover_min_opt 523 WRITE(numout,*) ' Maximum ocean coverage ', oce_cover_max_opt 524 WRITE(numout,*) ' nb of proc with coverage < 10 % ', inf10_opt 525 WRITE(numout,*) ' nb of proc with coverage 10 < nb < 30 % ', inf30_opt - inf10_opt 526 WRITE(numout,*) ' nb of proc with coverage 30 < nb < 50 % ', inf50_opt - inf30_opt 527 WRITE(numout,*) ' Number of computed points ', (npni_opt*npnj_opt-nland_opt)*jpimax_opt*jpjmax_opt 528 WRITE(numout,*) ' Overhead of computed points ', (npni_opt*npnj_opt-nland_opt)*jpimax_opt*jpjmax_opt-npiglo*npjglo 529 WRITE(numout,*) ' % sup (computed / global) ', ratio_opt 530 WRITE(numout,*) 531 ENDIF 532 CLOSE(numout) 533 ! 534 STOP 535 END PROGRAM mpp_optimize
Note: See TracChangeset
for help on using the changeset viewer.
