1 votes 1 votes in floating point representation we represent exponent as 2E-bias how is this bias value found? some places we use 8 , 16 , 64 etc A_i_$_h asked Nov 11, 2017 A_i_$_h 600 views answer comment Share Follow See all 5 Comments See all 5 5 Comments reply Shubhanshu commented Nov 11, 2017 reply Follow Share Excess value varies from question to question, But in IEEE 754 single and double precision, standard form is " -2|Exp|-1-1 " where |Exp| = number of bits allocated for Exponent part. As in single it is -127 and in double it is -1023. 1 votes 1 votes srestha commented Nov 11, 2017 reply Follow Share 16?? We doesnot use 16 anywhere right? 0 votes 0 votes A_i_$_h commented Nov 11, 2017 reply Follow Share @srestha to represent negative number in exponent we take 2's complement so it may give any value not only 16 @shubhanshu Suppose we store 0.100111 * 23 as 10 bits with mantissa 5 bits then how will u store E here? what will be the bias? 0 votes 0 votes srestha commented Nov 11, 2017 reply Follow Share any reference? 0 votes 0 votes A_i_$_h commented Nov 12, 2017 reply Follow Share @srestha i have it in my notes i don exactly remember where i referred it IF exponent is negative then how will u rep it then....?? if not taking 2'complement? 0 votes 0 votes Please log in or register to add a comment.