What is the largest number a data type can accurately store?

[bobbert]

Okay... I know that Decimal is accurate up to 29 digits and that Single and Double can store even larger numbers... what i want to know is: how much accuracy is lost using Single and Double? I think, because they use floating point, S and D become quite inaccurate *scratch* I'm not too sure...

if anyone could help it would be a huge help

thanks