measure pulse width 0.12 ns resolution
indicates that you want to measure a pulse not a running clock.

To achieve this you need at a minimum a clock of
1/resolution = 1/120p = 8.33 GigaHertz.

This will exculude any microcontroller I know of.

The chip (CPLD?) you need to use for this must have a clock at 10GHz. The best I can find from Xilinx is 500MHz.

This, as stated in the post, may be possible, but is waaay beyond the scope of this forum.

If, however, we are talking about a continuous clock, not a pulse, some solutions may come forward.

Please clarify.

Erik