Suppose in IEEE single precision, the width of the exponent field is 3, not 8, and the width of the fraction field is 5, not 23.
(a) What should the exponent bias be?
(b) What are the largest and smallest nonnegative normalized floating point numbers in this system?
(c) What is the machine epsilon of this system?© BrainMass Inc. brainmass.com March 4, 2021, 5:35 pm ad1c9bdddf
(a) exponent bias = 2^(3−1) - 1 = 3
(b) The smallest nonnegative ...
The solution determines exponent bias.