Scalability refers to an optimization engine’s ability to handle large problems.
For example, the process of developing a schedule for your company’s field service engineers – where you must evaluate large amounts of low-level detail such as request priorities, engineer schedules, time windows and resource sharing – may result in a massive optimization problem. At the same time, your problem’s size should not affect your system’s performance or your ability to use it to guide your daily operations.
Since Veeroute’s combinatorial optimization engine can scale and its algorithms are parallelizable, you can improve its performance by adding from one to an unlimited number of CPUs to your configuration. The following image shows the relationship between the number of CPUs in a configuration and the optimization engine’s performance:
Min configuration (1 CPU)
Each additional CPU will drastically improve the system’s performance. For example, if Veeroute uses 1 CPU then adding each additional processor will significantly accelerate the calculations
Number of CPUs
At some point, the increase in performance won’t be as large as it was in the first section. The characteristics of the problem you’re trying to solve will determine when this change occurs.
This fragment represents the stage at which adding CPUs will no longer improve the engine’s performance. The first part of this section represents the fastest possible configuration.