CS2011, A Term 1999 Prof. Sergio A. Alvarez IEEE 754 Floating Point Standard (single precision version) Uses 32 bits distributed as follows: 1 sign bit (0 for positive, 1 for negative) 8 bits for exponent in excess 127 notation; store exp + 127 instead of exp 23 bits for normalized mantissa; leading 1 not stored Example: IEEE 754 representation of 27.25d 27.25 d = 16 + 8 + 2 + 1 + .25 = 11011.01 b = 1.101101 x 2^4 exponent in excess 127 is 4 + 127 = 131 = 128 + 2 + 1 = 10000011 b mantissa without leading 1 is 101101 IEEE representation of 27.25 is therefore 0 10000011 101101 (17 0’s) = 41DA0000 h