Floating point is fundamentally a trade off between enumerable numbers (precision) and range between minimum/maximum numbers, it exists because fast operations

Floating point is fundamentally a trade off between enumerable numbers (precision) and range between minimum/maximum numbers, it exists because fast operations on numbers are not possible with arbitrary precision constructs (you can easily have CPU/GPU operations where floating point numbers fit in registers, arbitrary precision by its very nature is arbitrarily large). With many operations this trade off makes sense, however its critical to understand the limitations of the model.