CIDR and VLSM are basically opposites. With VLSM you are breaking the traditional networks claases into smaller subnets, the purpose of this is to save ip addresses . For example say you have a segment that only has 6 hosts on it , there is no need to assign a whole class C to this segment, this would be a waste of 248 Ip addresses . sh instead of using network address 192.168.10.0/24 you can use VLSM and assign the ip address of 192.168.10.0/29 , The second will have room for 6 hosts. Routing protocols that support VLSM carry the subnet mask information in routing updates(classless) , Routing protocols that do not (classful)do not carry subnet maslk info in routing updates and assume the major network boundries .
CIDR is mainly used for route summarization.
say you have a range of subnets that are all in the same AS that cover 192.168.0.0 192.168.1.0 192.168.2.0 192.168.3.0 , instead of advertising all of these networks you can cover them all with one network statement 192.168.0.0/22 . This reduces the size of the routing tables and decreases CPU and memory cycles on the routers . CIDR is mainly used in BGP , and classless routing protocols.
