The document discusses subnetting by borrowing bits from the host portion of a default class C or class B IP address to create subnets, with examples showing how to determine the number of bits to borrow to get a desired number of subnets, and how this partitioning yields a specific number of usable subnets and hosts per subnet along with their IP ranges and subnet mask. It provides step-by-step explanations for subnetting sample class C and class B networks, determining subnet and host bit locations and quantities.