Pings (ICMP echo requests) going to the switch's sc0 interface have to be processed by the NMP software running on the switch, which then sends an ICMP echo reply. We chose to have the CPU spend most of its time doing things that are important to keeping the switch performing well, so 7-8ms is
about the quickest response time I'd expect to see on the Cat4000. On a very busy switch, I'd expect longer response times, because the CPU is off doing other things.
When you ping *through* the switch, the ICMP echo request is just another data packet that is forwarded by the switch hardware. The latency you'd expect
to see in that case is the round-trip forwarding delay through the switch
(usually very short, on the order of microseconds), plus the latency of the IP
stack you're pinging, plus any other delay in the network that the ICMP echo
request and reply have to traverse. The following page should help as well
http://www.cisco.com/warp/public/473/82.html#ping
Hope this helps