Im trying to understand system architecture as it relates to backplane speeds and switching fabrics.
Some devices have switching fabrics while others dont. Example 6500 can be installed with a SFM increasing the system throughput to 256 Gbps from 32 Gbps Shares Bus. First question
The Shared Bus is on the backplane of the switch and all cards have access to the bus. Does the SCP determine where the frames go or as the name states its a shared bus and all line cards see the data?
When a SFM is installed only certain cards can utilize the SFM. So does the SFM make switching decisions or does the SFM provide a larger 256 Gbps bus to the SCP??
All of this stemmed from product comparison. The Juniper M320 supports 320 half Duplex. The Cisco 7600, with sup720, supports an aggregate of 720 Gbps. Is the 720 half or full duplex?
A colleague of mind says that its 720 half duplex and 360 full duplex. And the M320 is 320Gbps half duplex and 160 Gbps Full Duplex. That threw me for a loop. I would think it would be the other way around. He goes on to refer this as marketing math or Cisco math. Can someone please explain?
Sorry so long. Thanks for reading.