> what were those beastly boards?

The particular board I'm staring at is a T1/E1 interface board from a cisco=
-5800 access server.
It implements 12x channelized T1/E1 interfaces (360 channels!), so it has 1=
2 T1/E1 controllers, a bunch of FPGAs, a not-insignificant MIPS cpu, and in=
terface(s) to the system backplane(s).  Other boards in the system included=
 a main CPU, and a 120-channel modem card (obviously a rather dated board. =
 Dial-up; how quaint!)

http://necsweb.com/media/catalog/product/cache/1/image/9df78eab33525d08d6e5=
fb8d27136e95/D/S/DS58-12CE1.JPG


> the best bang for your buck in automation in PCB layout is in constraints=
 and rule sets, so the tool will tell you whether the trace you just routed

To be fair, I never watched anyone route a board of this class, nor have I =
used a PCB package of the class involved.  I get the distinct impression th=
at there are features that would seriously blur the distinction between aut=
o-routing and manual routing as defined by a simpler program like EAGLE (su=
rely they didn't lay out the 12x T1 front ends separately, one at a time.) =
  I also get the impression that a statement like "I lay out the critical n=
ets by hand and only let the autorouter handle the simple and easy stuff" s=
till leaves more being autorouted than any board floating around the open-s=
ource HW world=85

BillW


--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.