##Compile & build Instructions##
.../optimizationCode$ mkdir build
.../optimizationCode$ cd build
.../optimizationCode/build$ cmake -DCMAKE_BUILD_TYPE=RELEASE -DOptiX_INSTALL_DIR=/home/cs/software/sivelab/NVIDIA-OptiX-SDK-3.6.2-linux64 -DCUDA_TOOLKIT_ROOT_DIR=/home/cs/software/sivelab/cuda_6.0 ..
.../optimizationCode/build$ make
##Run instructions##
.../optimizationCode/build$ cd src
.../optimizationCode/build/src$ mpirun -np {no. of processes to be created} --machinefile {file containing the list of slaves} ./quic --optfile={path to optimization file} --loglevel={info, debug, error}
eg.
mpirun -np 10 -v --machinefile ../../inputs/machines.txt ./quic --optfile=../../inputs/optfiles/optTest8D_LSM_job_antlr_tiny.opt --loglevel=info
optimizationCode
-
docs (documentation)
-
fitness_lib (fitness function shared lib)
-
include (Contains symbolic links to include directories all libraries in src)
-
inputs (sample optfiles and hostsfile)
-
src (source code)--|
-
MPI_Framework (Code which needs to be implemented by all jobs)
-
mpi_gpuplume (GPU Plume code implementing MPI_framework)
-
opt_grammar (ANTLR grammar for reading optimization files)
-
qes_spf_mpi (TODO)
-
Utils (Utility code)
-
-
tests (Visit this if you have any questions)
-
thesisDoc(Thesis story)
-
CMakeLists.txt (Top level cmake file)
-
README.md (you are reading it right now!)