It allows each machine to have a globally unique IP address, without any form NAT. NAT breaks several protocols, and generally makes life more difficult. Also, if you want to use the same port for several machines (like 443) and you only have a single IP address then it makes life more difficult. For example, we have a single IPv4 address on a DSL tail. However, our entire network is IPv6 enabled using a 6to4 tunnel. We have several of our servers providing an https service. Thanks to IPv6 I can now easily browse to any of these servers from any of the remote locations that we have already IPv6 enabled. We are looking to actively start using IPv6 at many of our clients this year. We'll be using 6to4 tunnels again (because no one sells IPv6 transit ...). This means we can load a management agent on their machines, and have our server talk to those agents over a VPN. Normally we would have the grief of ensuring their private IP address range is different from all of our other client addresses ranges. With every machine having a globally unqiue IPv6 address this is no longer an issue. Of course, there is always an IPv4 RFC1918 clash, and we end up with the option of limited management, or a renumbering exercise. The sooner everyone adopts IPv6, and we can easily get native IPv6 transit the better. -----Original Message----- Still, I'm a little puzzled why a single IPv4 address (with lots of NAT behind it) is really any different from a single IPv6 prefix with lots of bits on the right hand side.