This is wrong, for 2 reasons:

• the precision supported is different.
• FP has some special values which cannot be represented, such as Infinity and NaN