Primetime crosstalk analysis difference

[mr_vasanth]

I do understand that cross talk analysis will be different for Asynchronous and logically exclusive clocks groups in Primetime SI. But unable to exactly understand what is timing window and how the PT calculates the timing window for async and exclusive clock group. Among Async and exclusive clock groups which one result in more pessimistic cross talk analysis ?

 

[sharath666]

Good Question. Really got me thinking...
The following note is from Solvnet.
"For all three relationships, timing paths between the clocks are suppressed. The relationships differ in how crosstalk is handled. If clocks are asynchronous, the clocks are assumed to have an infinite window relationship with each other. If clocks are physically exclusive, only one clock can act as an aggressor or victim during crosstalk analysis; interactions between the clocks are not explored. If clocks are logically exclusive, the crosstalk computation is computed normally (synchronously if the clocks are synchronous) even though the timing paths between the clocks are not reported. "

Looks like logically exclusive is the answer...

 

[mr_vasanth]

Thanks Sharath. Even I have read this statement in Solvnet. But I could not exactly understand the meaning of "infinite window relationship" for asynchronous clocks and how it affects cross talk analysis. I feel a Physical Design engineer would be the right person to answer this question.

 

[sharath666]

I think "infinite window" is the best case relationship between clocks; one where there will be minimum chances of crosstalk. But as you say, a phy des eng or an STA eng is best equipped to answer this question or at least elaborate on my ideas.