The actual implementation on a FPGA will use fixed point math with reduced precision. To be able to emulate this a bit better, introduce an option to cap the precision of the cluster position in global coordinates.
Tagging @jahreda, @brosser, @lipeles, @imaznas so they are aware.