By default internal rounding is turned on.

In this case, the smallest number you can store in a variable is 0.000000000001

Numbers are accurate to 12 decimal places to the right of the decimal point.

When working with very small or very large numbers, as in Scientific or Engineering applications, rounding must be turned off.

When numeric values exceed 999999999999999 they are stored in an exponential form which looses precision. Above 999999999999999 answers will only be approximations.

Big Decimal string arithmetic allows calculations with any number of digits, accurate to the last decimal place.