confusion

Question

Bias formula for floating point representation is 2^k-1

bias formula for IEEE floating point representation is 2^(k-1)-1   Is it right????

Comments

bias = (2^(k-1)) -1

Thats how we calculate bias

---

but when at some places where IEEE is not mentioned then first formula is used .....but at some place latter one is used?

---

@G Phalkey  yes correct 

Answer 1

we do not use 2^(k-1) as bias in IEEE representation due to some exception as in IEEE we have some special values when Exponent=255 so if u will use 128 as bias then we can get exponent upto 127+128= 255 which is already reserved for some special values hence we use 2^(k-1) - 1