At around forty to fifty zooms, the floating point numbers that Fractal eXtreme uses for calculating fractals, run out of precision and can not be used to zoom any further. At this point, most Fractal exploration programs stop. Those that continue on switch to using fixed point numbers. Fixed point number routines can be written to any level of precision, thus allowing users to zoom deep into fractals.
Whereas floating point numbers run out of accuracy at forty to fifty zooms, Fractal eXtreme's deep zoom fixed point routines allow zooming in to a depth of seven thousand two hundred zooms. This is equivalent to over two thousand digits of accuracy.
As the magnification increases, the calculation speed gradually decreases due to increased iterations per pixel, and the increasing complexity of the math.