How can limit 'Thread count' instead of using 'all available processors'?

Answered

Kamal Chowdhury

• July 16, 2019 11:15

I am using Gurobi to solve a fairly large MILP problem (19633 continuous and 8428 integer variables) written in Pyomo. When I run the model, Gurobi uses all available resources, for example, 12 out of 12 processors of a 6-cores CPU. I want Gurobi to use a specified number of processors, not the all available. I have tried by using SolverFactory("gurobi",Threads=1), but it did not work.

0

• 

  Eli Towle

  Gurobi Staff

  • July 16, 2019 21:28
  • Edited

  Hi Kamal,

  This can be done as follows:

  optimizer = SolverFactory('gurobi')
  optimizer.options['threads'] = 1

  For more information about passing options to solvers, see the Pyomo documentation on the topic. I hope this helps!

  Eli

  1
• 

  Kamal Chowdhury

  • July 17, 2019 04:39

  

  

  

  Hi Eli,

  Thank you for the nice simple solution!

  Best,

  Kamal

  0

Please sign in to leave a comment.