ICODE-ADC/lammps
4
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@12400 f3b2605a-c512-4ea7-a41b-209d697bcdaa

Browse Source
This commit is contained in:
sjplimp
2014-09-04 15:50:18 +00:00
parent af4909ab51
commit 3b20ca9f1a
1 changed files with 20 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
bench/KEPLER/README
View File

@ -16,30 +16,37 @@ openmpi/1.8.1/intel/13.1.SP1.106/cuda/6.0.37.
------------------------------------------------------------------------ ------------------------------------------------------------------------
You can of course build LAMMPS yourself with any of the accelerator You can, of course, build LAMMPS yourself with any of the accelerator
packages for your platform. packages for your platform.
The build.py script will build LAMMPS for the various accelerlator The build.py script will build LAMMPS for the various accelerlator
packages using the Makefile.* files in this dir, which you can edit if packages using the Makefile.* files in this dir, which you can edit if
necessary for your platform. You must set the "lmpdir" variable at necessary for your platform. You must set the "lmpdir" variable at
the top of build.py to the home directory of LAMMPS as installed on the top of build.py to the home directory of LAMMPS as installed on
your system. Then typing, for example, your system. Note that the build.py script hardcodes the arch setting
for the USER-CUDA package, which should be matched to the GPUs on your
system, e.g. sm_35 for Kepler GPUs. For the GPU package, this setting
is in the Makefile.gpu.* files, as is the CUDA_HOME variable which
should point to where NVIDIA Cuda software is installed on your
system.
Once the Makefiles are in place, then typing, for example,
python build.py cpu gpu python build.py cpu gpu
will build executables for the CPU (no accelerators), and 3 GPU will build executables for the CPU (no accelerators), and 3 variants
variants (double, mixed, single precision). See the list (double, mixed, single precision) of the GPU package. See the list of
of possible targets at the top of the build.py script. possible targets at the top of the build.py script.
Note that the build.py script will un-install all packages in LAMMPS, Note that the build.py script will un-install all packages in your
then only install the ones needed for the benchmark. The Makefile.* LAMMPS directory, then only install the ones needed for the benchmark.
files in this dir are copied into lammps/src/MAKE, as a dummy The Makefile.* files in this dir are copied into lammps/src/MAKE, as a
Makefile.foo, so they will not conflict with makefiles that may dummy Makefile.foo, so they will not conflict with makefiles that may
already be there. The build.py script also builds the auxiliary already be there. The build.py script also builds the auxiliary GPU
GPU and USER-CUDA library as needed. and USER-CUDA library as needed.
The various LAMMPS executables are copied into this directory LAMMPS executables that are generated by build.py are copied into this
when the build.py script finishes each build. directory when the script finishes each build.
------------------------------------------------------------------------ ------------------------------------------------------------------------