IPv6: Benefits, Adoption, and Enterprise Deployment

IPv6

Internet Protocol, Version 6 (IPv6) is the successor to IPv4 and one of the foundational protocols of the modern Internet. The first IPv6 specification, RFC 1883, was published in December 1995, and the protocol has been refined many times since. Today, IPv6 carries a large and growing share of Internet traffic, and enterprises are increasingly moving beyond dual-stack to IPv6-mostly and IPv6-only networks.

Last updated: November 9, 2026

Why IPv6?

The rapid expansion of the Internet and the growing number of connected devices created a critical need for more IP addresses. The 32-bit address space of IPv4 supports approximately 4.3 billion unique addresses. That seemed sufficient when IPv4 was standardized in the early 1980s, but with the explosive growth of the Internet, it became clear that a more scalable solution was necessary.

IPv4 Exhaustion

IPv4 address space is effectively exhausted. The central IANA free pool ran out in February 2011, and the Regional Internet Registries (RIRs) now have little or nothing left to allocate. What remains is handed out in small blocks, through waiting lists, or under restrictive policies. Organizations that need more IPv4 addresses must reclaim unused space or buy or lease addresses on the transfer market. The long-term solution is IPv6.

IPv6 Address Space

IPv6 uses a 128-bit address space, which yields roughly 340 undecillion unique addresses.

340,282,366,920,938,463,463,374,607,431,768,211,456

The standard IPv6 subnet is a /64, which on its own provides more than 18 quintillion addresses.

18,446,744,073,709,551,616

With such a vast pool of addresses, network engineers no longer need to worry about how many hosts will fit on a subnet.

Additional Benefits

Although the vast address space is the most significant improvement, IPv6 offers several other advantages over IPv4, including the following.

Simplified Addressing

There’s no need to conserve addresses with IPv6. Network engineers are freed from the constraints of small allocations and can design sites and networks around logical boundaries instead. Hosts can configure their own addresses using Stateless Address Autoconfiguration (SLAAC), or receive them from a DHCPv6 server.

Globally Unique Addresses

With ample address space, every host can have a globally unique address. IPv6 restores the end-to-end communication model of the original Internet and removes the need for Network Address Translation (NAT) for IPv6 traffic. Globally unique doesn’t mean exposed, however. A stateful firewall at the network edge remains essential, just as it is with IPv4.

No Broadcasts

IPv6 eliminates broadcast addressing and uses multicast and anycast addressing instead. This provides more efficient and controlled ways to communicate with specific groups of devices or to reach the nearest instance of a service.

Fixed Header Length

IPv6 uses a fixed 40-byte base header, with optional extension headers. This design simplifies and speeds up packet processing in routers and other network devices.

No Router Fragmentation

IPv6 routers don’t fragment packets. Fragmentation is handled only by the sending host, which relies on Path MTU Discovery to learn the largest packet size it can send. Path MTU Discovery depends on ICMPv6, so blocking ICMPv6 at firewalls breaks IPv6 connectivity. This is one of the most common mistakes in enterprise IPv6 deployments.

IPv6 Adoption Today

IPv6 adoption has grown steadily and recently passed an important milestone, as the following measurements show.

  • On March 28, 2026, more than 50% of users accessing Google did so over native IPv6 for the first time.
  • Other measurements, including APNIC and Cloudflare, put global adoption somewhat lower, at around 40% to 43%.
  • Adoption varies widely by country. France and India are above 70%, while several countries remain in the single digits.

In the United States, OMB Memorandum M-21-07 directs federal agencies to transition to IPv6-only, which is driving IPv6-only designs in government and among the vendors that serve it. Windows has supported IPv6 for many years and prefers it over IPv4 by default when both are available.

IPv6 in the Enterprise

Most enterprises begin with dual-stack networks that run IPv4 and IPv6 side by side. The next step is to reduce or remove IPv4 from the network while keeping IPv4-only applications and services working. The following topics cover that journey.

IPv6-Mostly and IPv6-Only Networks

On an IPv6-mostly network, clients that can operate without IPv4 use DHCP option 108 to signal that they prefer IPv6-only, while legacy devices still receive IPv4 addresses. IPv6-only clients reach IPv4-only destinations through NAT64 and DNS64.

464XLAT and Windows CLAT (WinCLAT)

Some applications use IPv4 literals or IPv4-only APIs and can’t work over NAT64 alone. 464XLAT solves this with a client-side translator (CLAT) that works with the provider-side translator (PLAT). Microsoft now includes a CLAT in Windows, known as Windows CLAT (WinCLAT).

Remote Access and IPv6

Remote access solutions must also work well when clients or servers are on IPv6 networks. These posts cover IPv6 considerations for Always On VPN and Absolute Secure Access.

How I Can Help

I help organizations plan and deploy IPv6 on Windows networks, from dual-stack readiness through IPv6-mostly and IPv6-only. That includes the following.

Additional Information

Internet Protocol, Version 6 (IPv6) Specification (RFC 8200)

IPv6 Stateless Address Autoconfiguration (RFC 4862)

IPv6-Only Preferred Option for DHCPv4 (RFC 8925)

Stateful NAT64 (RFC 6146)

464XLAT: Combination of Stateful and Stateless Translation (RFC 6877)

IPv6 Adoption Statistics (Google)

IPv6 Capable Rate by Country (APNIC)

Start a Conversation

Planning an IPv6 deployment or moving toward IPv6-only? Tell me about your environment, and I’ll follow up, typically within one business day.

← Back

Thank you for your response. ✨