Problem: simulations may not need to use uint64_t throughout; investigate uint32_t and scaling

[michael-okeefe]

Problem Description

As a hypothesis, we should be able to run faster if we were using uint32_t vs uint64_t for the flow data. This assumption should first be checked via measurement. If it is indeed the case, we may wish to investigate whether uint32_t can be used.

The switch to uint64_t was done out of an abundance of caution as flow magnitudes were approaching the maximum limit of uint32_t. However, this problem should be able to be circumvented through use of scaling. Currently, all flows are simulated in Watts. However, the scale could be determined automatically perhaps or we could temporarily change to uint64_t when doing efficiency conversion and the like.

Either way, we should investigate whether this change makes a difference or not and then go from there.