Major Description
There is a set on random source providers implemented in the project, here I'd like to suggest creating a benchmark to actually provide such an experiment.
Once having such test everybody will be able to compare and pick more suitable for use case.