Hello Jagdev,
if an MSFC daughter card is present the C6500 becomes capable of multi layer switching.
You can check the presence of an MSFC using the show module command.
To be noted in current C6500 running in Native IOS mode you actually connect to the route processor (that is the MSFC in hardware).
Multi layer switching allows the device to perform L2 switching and L3 switching ( = routing) at wire speed.
All this is done by using CEF tables that store all the information required to perform appropriate packet manipulation ( packet rewrite ).
The CEF table is not accessed directly but using the TCAM table that is an efficient hardware based table that uses as search keys parameters for example the IP destination address, the IP source address and provides an action that can be a pointer to an appropriate next-hop ( allowing also for PBR when needed).
The C6500 is capable of distributed forwarding for those linecards that have an intelligent sub-module called DFC, a copy of the CEF table is downloaded to each DFC-equipped linecards.
Linecards missing a DFC and with a CFC use the central CEF table that is stored in the PFC ( that is part of the supervisor).
see the C6500 architecture white paper for a complete explanation
http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps708/prod_white_paper0900aecd80673385.html
the TCAM tables have a finite size that must be taken in account and that influences the choice of the type of supervisor and of the type of DFCs to be used in a C6500.
So to make an example to host a full internet table you need a Sup720 3BXL or Sup720 3CXL or other -XL for more modern supervisors as the XL indicates bigger TCAM tables.
For a campus network where only few thousands routes are present in the IP routing table normal supervisors and DFCs can be used that have smaller TCAM tables.
Exceeding the size of the TCAM table would mean some traffic process switched with high performance degradation as the main cpu usage would go very high.
Hope to help
Giuseppe
