The first roll does not matter.
The second has a five in six (0.833) chance of �success� (i.e. not matching the first die).
The third has a four in six (0.666) chance of success (i.e. not matching either of the previous two).
The fourth has 0.5 chance of success.
The fifth 00.333.
The sixth 0.167.
For all to be different each roll must be �successful� so these odds must be multiplied together and they amount to 120/7776 or 0.015432. Which is 1 in 64.8.
I do not believe the number 46,656 is relevant to this calculation.
The reason why the odds against six different numbers are considerably less than six the same is that the earlier dice have quite a high chance of success (five in six, four in six, etc.) whereas to match all six each die only has a one in six chance of success.
It does not matter whether the experiment consists of one die being rolled six times or six dice being rolled simultaneously.