Fixed vs floating point representation doubts

Question

I have an exam tomorrow for which I need to know about both fixed and floating point representations. I think I've understood the basic idea behind both, but when it comes to comparing their features I'm not quite sure about the details. I'll note down what I feel is correct from the understanding that I have about both, and would greatly appreciate it if someone could confirm whether it's correct or point out what's wrong.

Fixed point -

a. Faster than floating point implementation

b. Can represent any value accurately within its range

c. Allows simple multiplication by 2

Floating point -

a. Provides best resolution (I'm assuming resolution means precision)

b. Copes with a wide range of numbers

c. Can't represent some values with exact accuracy in its range

d. Implementation is slightly more complicated

Thank you.

Answer 1

Fixed point -

a. Faster than floating point implementation
- TRUE/FALSE - can be faster or slower depending on hardware

b. Can represent any value accurately within its range
- FALSE

c. Allows simple multiplication by 2
- "simple" compared to what ? For floating point you add 1 to the exponent, for fixed point you either do an integer multiply or a shift left. I don't see any significant difference in complexity.

Floating point -

a. Provides best resolution (I'm assuming resolution means precision)
- partially TRUE - but "best" compared to what ? Do you just mean better than fixed point ?

b. Copes with a wide range of numbers
- TRUE

c. Can't represent some values with exact accuracy in its range
- TRUE - but can't represent the vast majority of values with exact accuracy

d. Implementation is slightly more complicated
- TRUE - but again, "slightly more complicated" than what ? Floating point requires a lot more logic (i.e transistors/gates/silicon) than fixed point, if that's what you mean ?