The timing path is not consistent between Encounter and PrimeTime

[kunfun2002]

I have one timing path and its start point and end point are all the same in Encounter and Primetime.

But the problem is that the inside logic which trace by Encounter and Primetime are different

Encounter show timing path from one A register/Q to B register/D
But primetime timing report show from one A register/Q to one RAM/A pin and go out from RAM/O pin and reach to B register/Q
The problem is that primetime report's timing path has violation and it is true path.
But the Encounter cannot see it.
How should I fix it?


Translate to Chinese for more people can understand my question:
我有一條 timing path的起始點和終點 在Encounter和PrimeTime都是一樣
問題是裡面走的路徑和邏輯 在Encounter和PrimeTime走的路徑不一樣
雖然最後timing report報出來是相同的起始點和終點

其中Encounter報的timing path 是從一個register/Q到 一個register/D

但是primetime timing report卻是報從一個register/Q 先經過一個RAM的A pin再從RAM的O pin出來 到re偶gister/D

問題是primetime報出來的timing path的timing path有violation而且是true path

但是Encounter一直看不到這條path

請問該如何解?

 

[rca]

First you need to report the same path, I means through the same logic to understand what Encouter reports.
After to "debug":
1- report the same path in both tool, without any derate factor on cells or nets and without SI. The result must be the same at ps.
a- The parasitic extraction should be very similar, if not adjust the cap&res factor in Encounter.
b- check all transitions(slew)/cap reported by both tool.
2- Add the SI mode, in both tool, the result should be closed.