Contents of /MITgcm/doc/README

$Header: /u/gcmpack/models/MITgcmUV/doc/README,v 1.12 1998/07/16 16:58:32 cnh Exp $


MITgcmUV Getting Started
========================

o Introduction

  This note is a guide to using the MIT General Circulation Model Ultra Verstaile
  implementation, MITgmcUV. MITgcmUV is a Fortran code that implements the 
  algorithm described in Marshall et. al. 1997, Hill, Adcroft, ...
   The MITgcmUV implementation is designed to work efficiently on all classes
  of computer platforms. It can be used in both a single processor mode
  and a parallel processor mode. Parallel processing can be either multi-threaded
  shared memory such as that found on CRAY T90 machines or it can be multi-process
  distributed memory. A set of "execution enviroment" support routines are 
  used to allow the same numerical code to run on top of a single-process, multi-threaded
  or distributed multi-process configuration.

o Installing
  To setup the model on a particular computer the code tree must be created
  and appropriate compile and run scripts set up. For some platforms
  the necessary scripts are included in the release - in this case follow
  the steps below:

  1. Extract MITgcmUV from the downloadable archive
     tar -xvf checkpoint12.tar 

  2. Create platform specific make file
     For example on a Digital UNIX machine the script "genmake.dec" can
     be used as shown below

     cd bin
     ../tools/genmake 
     cp Makefile.alpha Makefile                  ( On Alpha machine)

  3. Now create header file dependency entries
     make depend

  4. Compile code
     make

  5. Copy input files
     cp ../verification/exp2/[a-z]* ../verification/exp2/*bin .
  
  6. Run baseline test case
     setenv PARALLEL 1
     dmpirun -np 2 ../exe/mitgcmuv        ( Under Digital UNIX   )
     mpirun.p4shmem ../exe/mitgcmuv -np 2 ( Under Solaris + mpich)
                                                      

     This runs a 4 degree global ocean climatological simulation.
     By default this code is set to use two processors splitting
     the model domain along the equator. Textual output is written
     to files STDOUT.* and STDERR.* with one file for each process.
     Model fileds are written to files suffixed .data and .meta
     These files are written on a per process basis. The .meta
     file indicates the location and shape of the subdomain in
     each .data file.
    
     This 4 degree global ocean climatological simulation is the baseline
     configuration for the MITgcmUV code. The change files that 
     convert the model to a different configuration in the 
     verification directory all assume that the model is configured for
     the baseline case and change the model code accordingly.

o Running

  - Input and output files

    Required files
    ==============
    The model is configured to look for two files with fixed names.
    These files are called
     "eedata" and "data".
    The file eedata contains "execution environment" data. At present
    this consists of a specification of the number of threads to
    use in X and Y under multithreaded execution.

  - Serial execution

  - Parallel execution. Threads
    nSx, nSy
    setenv PARALLEL n
    nTx=2, nTy=2

  - Parallel execution. MPI
    mPx, nPy
    dmpirun

  - Parallel execution. Hybrid

o Cutomising the code

  Model structure
  ===============
  The "numerical" model is contained within a execution 
  environment support wrapper. This wrapper is designed 
  to provide a general framework for grid-point models.
  MITgcmUV is a specific numerical model that uses the
  framework.
   Under this structure the model is split into 
  execution environment support code and conventional
  numerical model code. The execution environment
  support code is held under the "eesupp" directory.
  The grid point model code is held under the 
  "model" directory. 
   Code execution actually starts in the eesupp 
  routines and not in the model routines. For this
  reason the top level main.F is in the eesupp/src
  directory. End-users should not need to worry about
  this level. The top-level routine for the numerical
  part of the code is in model/src/the_model_main.F.


o References
  Web sites - HP
  for doc     Digital
              SGI
              Sun
              Linux threads
              CRAY multitasking
              PPT notes
1	$Header: /u/gcmpack/models/MITgcmUV/doc/README,v 1.12 1998/07/16 16:58:32 cnh Exp $
2
3
4	MITgcmUV Getting Started
5	========================
6
7	o Introduction
8
9	This note is a guide to using the MIT General Circulation Model Ultra Verstaile
10	implementation, MITgmcUV. MITgcmUV is a Fortran code that implements the
11	algorithm described in Marshall et. al. 1997, Hill, Adcroft, ...
12	The MITgcmUV implementation is designed to work efficiently on all classes
13	of computer platforms. It can be used in both a single processor mode
14	and a parallel processor mode. Parallel processing can be either multi-threaded
15	shared memory such as that found on CRAY T90 machines or it can be multi-process
16	distributed memory. A set of "execution enviroment" support routines are
17	used to allow the same numerical code to run on top of a single-process, multi-threaded
18	or distributed multi-process configuration.
19
20	o Installing
21	To setup the model on a particular computer the code tree must be created
22	and appropriate compile and run scripts set up. For some platforms
23	the necessary scripts are included in the release - in this case follow
24	the steps below:
25
26	1. Extract MITgcmUV from the downloadable archive
27	tar -xvf checkpoint12.tar
28
29	2. Create platform specific make file
30	For example on a Digital UNIX machine the script "genmake.dec" can
31	be used as shown below
32
33	cd bin
34	../tools/genmake
35	cp Makefile.alpha Makefile ( On Alpha machine)
36
37	3. Now create header file dependency entries
38	make depend
39
40	4. Compile code
41	make
42
43	5. Copy input files
44	cp ../verification/exp2/[a-z]* ../verification/exp2/*bin .
45
46	6. Run baseline test case
47	setenv PARALLEL 1
48	dmpirun -np 2 ../exe/mitgcmuv ( Under Digital UNIX )
49	mpirun.p4shmem ../exe/mitgcmuv -np 2 ( Under Solaris + mpich)
50
51
52	This runs a 4 degree global ocean climatological simulation.
53	By default this code is set to use two processors splitting
54	the model domain along the equator. Textual output is written
55	to files STDOUT.* and STDERR.* with one file for each process.
56	Model fileds are written to files suffixed .data and .meta
57	These files are written on a per process basis. The .meta
58	file indicates the location and shape of the subdomain in
59	each .data file.
60
61	This 4 degree global ocean climatological simulation is the baseline
62	configuration for the MITgcmUV code. The change files that
63	convert the model to a different configuration in the
64	verification directory all assume that the model is configured for
65	the baseline case and change the model code accordingly.
66
67	o Running
68
69	- Input and output files
70
71	Required files
72	==============
73	The model is configured to look for two files with fixed names.
74	These files are called
75	"eedata" and "data".
76	The file eedata contains "execution environment" data. At present
77	this consists of a specification of the number of threads to
78	use in X and Y under multithreaded execution.
79
80	- Serial execution
81
82	- Parallel execution. Threads
83	nSx, nSy
84	setenv PARALLEL n
85	nTx=2, nTy=2
86
87	- Parallel execution. MPI
88	mPx, nPy
89	dmpirun
90
91	- Parallel execution. Hybrid
92
93	o Cutomising the code
94
95	Model structure
96	===============
97	The "numerical" model is contained within a execution
98	environment support wrapper. This wrapper is designed
99	to provide a general framework for grid-point models.
100	MITgcmUV is a specific numerical model that uses the
101	framework.
102	Under this structure the model is split into
103	execution environment support code and conventional
104	numerical model code. The execution environment
105	support code is held under the "eesupp" directory.
106	The grid point model code is held under the
107	"model" directory.
108	Code execution actually starts in the eesupp
109	routines and not in the model routines. For this
110	reason the top level main.F is in the eesupp/src
111	directory. End-users should not need to worry about
112	this level. The top-level routine for the numerical
113	part of the code is in model/src/the_model_main.F.
114
115
116	o References
117	Web sites - HP
118	for doc Digital
119	SGI
120	Sun
121	Linux threads
122	CRAY multitasking
123	PPT notes