Floating-point numbers are numbers which use a decimal point to describe their values, (such as 17.5). The math co-processor can process such numbers more efficiently which is why a computer with a co-processor can perform math-intensive functions faster than without one.
