No, if you mean "dx" here in the usual sense of a differential, it is NOT 0.
There are a number of different ways of interpreting "dx" depending on how you are handling Calculus itself. If you are using "Non-standard Analysis" (Calculus becomes very simple in "Non-standard Analysis" but it requires some very deep logic theory to show that it is valid), then "dx" and all differentials are "infinitesmals". If you use the more "standard" limit approach to Calculus, "dx" is purely a "symbol" that has no meaning by itself but can be connected to the ordinary derivative by "dy= f'(x)dx". In either case, it is not 0 (unless, of course, x itself represents a constant function).
(Hah! I finally beat someone by one minute!)
(Usually, I am the one who posts one or two minutes after another.)