Wolfram Language & System 11.0 (2016)|Legacy Documentation
When you use //N to get a numerical result, the Wolfram Language does what a standard calculator would do: it gives you a result to a fixed number of significant figures. You can also tell the Wolfram Language exactly how many significant figures to keep in a particular calculation. This allows you to get numerical results in the Wolfram Language to any degree of precision.
|expr//N or N[expr]||approximate numerical value of expr|
|N[expr,n]||numerical value of expr calculated with n‐digit precision|
Doing any kind of numerical calculation can introduce small roundoff errors into your results. When you increase the numerical precision, these errors typically become correspondingly smaller. Making sure that you get the same answer when you increase numerical precision is often a good way to check your results.