Previously, we noted how important it is to create balanced workloads, where the compute time for each task is equal.
When the nature of the problem is such that there are many more tasks available than workers and each task is truly independent, then a task farm is a simple parallel processing scheme that ensures 100% utilization of workers by the master feeding the next available task to the next free worker, as depicted in the following diagram:
In this case, it does not matter that each task varies as to the amount of compute it requires as there is no intertask dependency during the compute phase (at least).