why does floating point arithmetic consums more power than fixed point number arithmetic?

As I mentioned in the title, why floating points adds|muls|div consumes more power than fixed points adds|muls|div?

For instance, what does the flowing code exactly do on the CPU chip?

float a = 1, b = 2, c = 3;  c = a*b + c;  

Is the above code consumes more power than the following one? Why?

int32_t a = 1, b = 2, c = 3;  c = a*b + c;  

https://stackoverflow.com/questions/65572285/why-does-floating-point-arithmetic-consums-more-power-than-fixed-point-number-ar January 05, 2021 at 10:06AM