Dividing an int by zero, will throw an exception, but a float won\'t - at least in Java. Why does a float have additional NaN info, while an int type doesn\'t?
A clear Explanation about float arithmetic is given here
http://www.artima.com/underthehood/floatingP.html