First Convert 10.25 to binary format in this way
(10.25)10=(1010.01)2
For IEEE 754 format we have to normalize the number (decimal point following a leading 1): 1.01001*23
Now, the leading 1 here is omitted while storing as it is implied as long as exponent is non zero (normalized representation)
here for Single precision Biased exponent= E'=E+127 (127 is bias)
E'=130=(10000010)2
So for 32 bits:
Sign bit(1 bit) |
exponent bits(8bits) |
mantissa(23bits) |
0(for +ve sign) |
10000010 |
01001000000000000000000 |
convert it into hexadecimal format:41240000