FloatingPointRepresentation

# A pattern of all zeros in the fraction eld of a

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ortly. A pattern of all zeros in the fraction eld of a normalized number represents the signi cand 1.0, not 0.0. Zero is not the only special number for which the IEEE standard has a special representation. Another special number, not used on older machines but very useful, is the number 1. This allows the possibility of dividing a nonzero number by 0 and storing a sensible mathematical result, namely 1, instead of terminating with an over ow message. This turns out to be 5 Table 1: IEEE Single Precision a1a2a3 : : :a8 b1b2b3 : : :b23 If exponent bitstring a1 : : :a8 is (00000000)2 = (0)10 (00000001)2 = (1)10 (00000010)2 = (2)10 (00000011)2 = (3)10 # (01111111)2 = (127)10 (10000000)2 = (128)10 # (11111100)2 = (252)10 (11111101)2 = (253)10 (11111110)2 = (254)10 (11111111)2 = (255)10 Then numerical value represented is (0:b1b2 b3 : : :b23)2 2;126 (1:b1b2 b3 : : :b23)2 2;126 (1:b1b2 b3 : : :b23)2 2;125 (1:b1b2 b3 : : :b23)2 2;124 # (1:b1b2b3 : : :b23)2 20 (1:b1b2b3 : : :b23)2 21 # (1:b1b2b3 : : :b23)2 2125 (1:b1b2b3 : : :b23)2 2126 (1:b1b2b3 : : :b23)2 2127 1 if b1 = : : : = b23 = 0, NaN otherwise very useful, as we shall see later, although one must be careful about what is meant by such a result. On...
View Full Document

## This note was uploaded on 02/12/2014 for the course MATH 4800 taught by Professor Lie during the Spring '09 term at Rensselaer Polytechnic Institute.

Ask a homework question - tutors are online