Usage
CLB/MODEL_NAME/main case_path
Where MODEL_NAME
is the name of the desired model(e.g.d2q9
) and case_path
is the location of case file(.xml), relative to the TCLB folder.
For example, to run karman.xml
case, located in subfolder example\flow\2d
, using d2q9
model:
CLB/d2q9/main example/flow/2d/karman.xml
CPU-based usage
By default the code will run on GPU, which requires a NVIDIA GPU with CUDA support. In order to run it on CPU, the code must be compiled using --disable-cuda
option. In order to run the calculations on CPU in parallel, one need to start calculations using mpirun
option:
mpirun -np X CLB/MODEL_NAME/main case_path
Where X
is number of threads that will be created. Usually max number of threads is equal to number of cores of a processor, with the exception of CPUs with multi-threading.