set_clock_uncertainty does not affect IO timing analysis

[mongrainjr]

set_clock_uncertainity

I use: set_clock_uncertainty 0.4 [get_ports CLK_PIN]

in DC and Primetime. When I write the timing report I see the 0.4ns is taken into account in the Flop-to-flop analysis. However, It is not taken into acount in the Flop-to-PAD analysis.

The set_clock_uncertainty is use to emulate the skew but also the Jitter on the clock. In a Flop-to-PAD analysis we need to take into account Jitter. Is there a way to emulate teh Jitter for Flop-to-PAD analysis ?

 

[coolrak]

set_clock_uncertainty

set_clock_uncertainity is for skew. As per my knowledge clock jitter effect will be

specified in set_clock_latency.

 

[johnz]

io timing analysis

you must take account clock insertion delay for analysis of IO timing.

to over-constrain boundary logic, margin should be added to input delay and output delay, clock uncertainty don't represent margin.

 

[stocking]

set_clock_uncertainty jitter

I think the uncertainty means jitter, the delay from Flop to PAD is calculated in output_delay.