IPv6 Is 30 Years Old—Why It Still Hasn’t Won Over the Internet

IPv6 turned thirty this year, yet the protocol has not replaced IPv4 as the dominant Internet addressing scheme. Its most obvious advantage is the expansion from a 32‑bit space (about 4.3 billion addresses) to a 128‑bit space (approximately 3.4 × 10³⁸ addresses), a capacity that could assign an address to every grain of sand on Earth.

In the early 1990s, a small group of engineers calculated that the 4.3 billion IPv4 addresses would soon be insufficient for the growing number of devices, predicting a crisis that would force a rapid global transition. RFC 1883, published in December 1995, formally defined IPv6 with the promise of solving the address shortage.

The expected migration never materialised. In 2011 the IANA IPv4 pool was exhausted, and headlines warned of an imminent "Internet apocalypse". However, adoption statistics from Google, APNIC and Cloudflare show that less than half of global users now use IPv6, and the growth curve resembles a tree in a desert—slow and never reaching a forest.

“IPv6 is an extremely conservative protocol that changes as little as possible. It is a classic case of committee‑driven bad design.” – Geoff Huston, APNIC chief scientist

The protocol’s functional enhancements over IPv4 are minimal. Features such as security extensions, automatic configuration, QoS and multicast were later added to IPv4, so the perceived "next‑generation" benefits were largely absent.

Deploying IPv6 also incurs significant operational costs: new router firmware, firewall rule changes, staff training, script rewrites and extensive compatibility testing. Gartner analyst Andrew Lerner even observed that some organisations disable IPv6 to improve performance, indicating that in certain scenarios IPv6 can be slower than IPv4.

Conversely, Network Address Translation (NAT) emerged as a cheap, effective patch for IPv4 address exhaustion. By allowing many private devices to share a single public address, NAT extended the usable life of IPv4 and unintentionally provided a security barrier, as external scanners see only one IP address.

“These solutions are relatively easy to deploy, align with existing operational expertise, and avoid large‑scale infrastructure changes. For most operators, continuing with NAT on IPv4 is far simpler than migrating to IPv6.” – Alvaro Vives, RIPE NCC

While IPv6 struggled to win on its own merits, it quietly became the backbone for the massive influx of new devices—mobile phones, IoT sensors, cloud instances—because its vast address pool enables painless scaling for infrastructure providers.

At the same time the Internet’s architecture is shifting from an address‑centric model to a name‑ and service‑centric model. Protocols such as QUIC (RFC 9000) use DNS names rather than IP addresses for transport, reducing the importance of the underlying IP version.

In the final analysis, IPv4 and IPv6 will coexist for the foreseeable future. IPv4 persists because it is cheap, compatible with legacy equipment, and requires no changes. IPv6 survives as a low‑cost, abundant address source for the "new world" of mobile, cloud and IoT, acting as an invisible foundation beneath everyday services.