It took 28 years, but IPv6 usage on Google's servers has surpassed IPv4 for the first time in history — and it's far more interesting than it sounds.
-
IP addresses are the phone numbers of the internet.
Every device and server on the internet has an IP address. The reason we don't see these numbers when we browse to a website is that a DNS server quietly translates the web address we typed into the corresponding IP address behind the scenes — just like a contact name in your phone.
The IPv4 protocol launched around 1980, relying on 32-bit addresses capable of representing 4.3 billion unique addresses (of which only 3.7 billion are actually usable).
Back then, that sounded like plenty. Nobody imagined the wild world we live in today, where every home is packed with dozens of smart devices, each needing its own address.
Between 2011 and 2019, IPv4 addresses ran out, and to this day, if you want to lease one for your server, you'll have to pay per use.
One solution to this problem is NAT. With this approach, each local network — like the one in your home — has just a single IP address, assigned to the main router. All devices within that network use generic internal addresses, eliminating the need to allocate a separate IP to every device.
-
The IPv6 protocol, dating from 1998, uses a 128-bit hexadecimal representation. It can represent a truly staggering 340,282,366,920,938,463,463,374,607,431,768,211,456 addresses — enough to assign several billion addresses to every grain of sand on Earth.
What do you do with that many addresses? First, you burn half of them.
The first half is managed by the router and represents the network address; only the second half remains to identify the specific device within the network.
Under this protocol, each device generates its own precise device address using a complex algorithm based on the physical address of the network card and additional variables, and this address refreshes automatically from time to time.
-
Why should any of this matter to us?
This protocol makes the internet significantly faster and more secure.
Because every device generates its own address in an instant — without going through a router's allocation process and mapping tables — communication becomes more efficient and stable, and failures are reduced.
Because that address changes randomly, it blocks tracking based on a user's IP address. And on top of all that, addresses exist in abundance and are free.
Despite these clear advantages, a lack of supporting devices prevented widespread adoption of the protocol, which in turn discouraged the necessary infrastructure investment — a chicken-and-egg cycle that kept repeating.
Only now, 28 years after its introduction, has Google announced that more than half of the traffic on its servers relies on IPv6. From here, the trend will only accelerate on the road to a faster, more modern internet that we'll all benefit from.
--
👋 Hi, I'm Shlomo Strauss — follow me for more content on science and technology.