size and range of basic data type depency on the architecture of computer

is the size and range of any data type,suppose "float", depends on the architecture of the processor of the CPU?

for 16 bit,32 bit and 64 bit: the size and range of float are different ? why?