I find the recommended methodology quite appropriate (in fact it is amazing like
the coupling file used in XTK, I guess function does define form). There is one situation
we want to make sure and plan for and that is where there are a hundred pins each of which
couples to its immediate neighbors but doesn't really couple to pins a few pins away. (I think
the propose format covers this syntactically but we want to be ready to simulate it also).
1
2
3
4
5
6
7
8
9
so
123 couple
234 couple
345 couple
567 couple
etc.
This is important because it allows the database to be partitioned into simulation segments that overlap but are
smaller than simulating the entire pin coupling set. (which is essential for efficient simulation).
jon
Received on Thu May 4 17:51:35 1995
This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:29 PDT