Hello,
I am trying to use pybnb for solving a combinatorial optimization on 30 boolean variables.
The dimensional of the problem is quite huge, so I am using a computing cluster with MPI.
I found out that using multiple nodes, only one nodes increases the memory usage, so I deduced the other nodes are only used for computation.
That represents a bottlekneck beacause when the problem is very large, it should be better to span the memory usage on multiple nodes.
Please, let me know how this drawback can be solved.
Thanks for help.
Regards.
Ivan
Hello,
I am trying to use pybnb for solving a combinatorial optimization on 30 boolean variables.
The dimensional of the problem is quite huge, so I am using a computing cluster with MPI.
I found out that using multiple nodes, only one nodes increases the memory usage, so I deduced the other nodes are only used for computation.
That represents a bottlekneck beacause when the problem is very large, it should be better to span the memory usage on multiple nodes.
Please, let me know how this drawback can be solved.
Thanks for help.
Regards.
Ivan