RE: PCI clock wire routing

["John R Pierce" <pierce@scruznet.com>]

Francois Lepinette wrote...

>Recently I've seen a strange way of routing the PCI clock
>track from the PCI connector to the PCI interface chip.

>It was routed on a IBM PCI to SCSI card bridge as follow :
(serpentine routing graphics deleted...)

In fact, this is quite common.  As several folks have pointed out, the 
CLK signal has a MIN and MAX length... Looking at a couple of PCI
video cards I have, they ALL have this sort of zig zag routing of CLK.

Heh. You think this is kind of strange, you should see the routing 
layout required by "RDRAM" (aka RamBus)...

-jrp

A0