All IP addresses were divided into classes when they were first devised. The idea was that Class A addresses were allocated to large companies. Class B to medium sized companies and Class C to any other company who needed an IP address. Class D is for multicasting such as streaming media and Class E is reserved for experimental use.
You can recognise a Class A address because the first octet is a number from 1 to 126 inclusive. So any of the below IP addresses are class A.
Network devices recognise a Class A address as such because the numbers are checked in binary and the first number in the first octet always begins with the number 0. Check the binary section for more info on that.
Class A addresses lock the first octet for network numbers. The remaining three octets can be used for host numbers. It works like this:
In the above example 20 is the network number and 23.2.4 is a host on that network.
Please note that Class A, B and C addresses are only called such for historical purposes now and you may find that your company could be allocated a Class A address or even one for your home IP address for your broadband connection. VLSM means that we don’t have to concern ourselves with Classes of address anymore.