OR/MS Forecast: Sunny days for the Cloud

Cloud computing is everywhere this year. A lot of researchers are proposed parallel algorithms for many applications. The Cloud is also quite the hype with software vendors. Gurobi, CPLEX and Sulumn, among others, have announced or are already shipping distributed computing algorithms for solving mathematical programming problems in the cloud. You can thus build your model on your laptop and then distribute the workload of solving the model on one or more servers.

There are essentially two advantages for cloud-based optimization:

  1. The ability to use optimization without the usual fixed costs (expensive hardware, software licenses, etc.)
  2. The ability to use more processing power in the hope of solving larger or more complex models.

The first point is easy to make. Pay-per-use is excellent for teams that have either limited use for optimization or need extra processing power for peak  periods.  More options for users and less programming efforts to develop decision support systems allows for a very positive future; I expect to see even more use of optimization-based tools in the future, both in research and practice.

The second point is less obvious. For instance, we know that using more computational resources does not necessarily lead to reduction in solving times, at least for mixed-integer programs (there are posts showing that for CPLEX and Gurobi, for those interested). However, the widespread availability of state-of-the-art cloud computing tools will help accelerate research on new or improved optimization algorithms designed for the Cloud. Expect to see more clouds on the horizon in the near future!

Speak Your Mind

*