Use multiple machines to find parameter settings that improve performance

Gurobi includes a parameter tuning tool that automatically searches for parameter settings that improve optimization performance on a model or a set of models. This tuning tool has proven to be both quite effective and extremely popular. One downside of the tuning tool is the time it can take. Automated tuning performs many solves on your model, each using different parameter settings. By default, this search for improved parameter settings takes roughly 10 times as long as solving the model (and it is often beneficial to let the tuning tool run for much longer). The distributed tuning tool uses multiple, independent machines to perform tuning, thus allowing you to explore more candidate parameter settings in the same amount of time. Internal benchmark tests show mean performance improvements of over 2X when using tuned settings, as compared with default parameter settings.


When to use distributed tuning

Distributed tuning is useful when you need to find settings that maximize the performance of a single machine solve.