Total it is 64 bits for a 8 byte number (SAS default)

    Sign bit: 1 bit
    Exponent width: 11 bits
    Significand precision: 53 bits (52 explicitly stored)



That is how precision digits it can go.

data _null_;
do i= 50 to 55;
  x=2**(-i);
  put i= x= 19.16;
  end;
run;

Thanks. Your code is to generate the new data with 16 decimal. My purpose is to input and display the existing data with 16 decimal. Please let me know how to do.

e.g.
data test;
input a 19.16;
cards;
0.11111111111113112
;
run;
proc print data = test;
var a;
format a 19.16;
run;

The results is still
Obs a
1 0.1111111111111300, not 0.11111111111113112