IPv4 vs IPv6
The differences between the two IP versions that actually matter in day-to-day network engineering — addressing, configuration, NAT, headers and more, without vendor marketing.
Published
Side-by-side comparison
| Attribute | IPv4 | IPv6 |
|---|---|---|
| Address size | 32 bits | 128 bits |
| Notation | 192.168.1.1 (four dotted-decimal octets) | 2001:db8::1 (eight hex groups, compressible) |
| Total addresses | 4,294,967,296 | ≈3.4 × 1038 |
| Configuration | Manual/static or DHCP | SLAAC (stateless autoconfiguration) or DHCPv6 |
| NAT | Ubiquitous, largely because public addresses are scarce | Not required by design; end-to-end addressing is the norm |
| Header | Variable length (20–60 bytes), includes a checksum | Fixed 40-byte header, no IP-layer checksum |
| Broadcast | Supported | No broadcast; multicast instead, including efficient solicited-node multicast for neighbor discovery |
| Fragmentation | By the sending host and intermediate routers | By the sending host only; routers rely on Path MTU Discovery |
In raw capacity terms, IPv6 offers about 7.9 × 1028 times as many addresses as IPv4. The two notations are not interchangeable: an IPv4 address never contains letters or colons, and IPv6 uses prefix lengths (/64) rather than dotted masks. The leaner v6 packet format is codified in RFC 8200, which standardized Internet Protocol version 6 in its modern form.
Why IPv6 exists at all
IPv4's 4,294,967,296 addresses once seemed inexhaustible. They were not: IANA distributed its last central blocks in 2011, and every regional registry has since exhausted its own free pool. The short-term patches — NAT, carrier-grade NAT, address leasing markets — work, but they add complexity and break the original end-to-end model of the internet. IPv6's 128-bit space removes the scarcity that made those workarounds necessary.
Where you still need IPv4 today
- Legacy systems. Old appliances, industrial controllers, printers and embedded devices often have no IPv6 stack at all.
- Enterprise software assumptions. Firewall rules, ACLs, monitoring dashboards and licensing schemes are frequently written against IPv4 syntax first — if at all.
- Carrier environments. Mobile and consumer ISPs commonly run IPv4 over carrier-grade NAT (the 100.64.0.0/10 range), so customer-facing troubleshooting still starts with v4.
- Interoperability. Until every peer, resolver and path speaks v6, dual-stack networks must keep full IPv4 routing, DNS and security tooling alive.
The practical takeaway: learning IPv4 subnetting is not optional, even on networks with healthy IPv6 deployment. Our IPv4 subnet calculator and VLSM calculator cover that side.
The dual-stack reality
Most transitioned networks run both protocols side by side: every interface gets an IPv4 address and one or more IPv6 addresses, and applications use whichever the DNS answer suggests (A records for v4, AAAA records for v6). That means double the addressing plan, double the firewall policy surface and double the troubleshooting matrix — but it is how real networks operate during the long coexistence period.
When you do get a native IPv6 allocation, the planning math changes shape: you typically receive a /48 or /56 and slice it into /64 LANs. The layout of the global unicast space those allocations come from is tracked in the IANA IPv6 address space registry. Work out the details with the IPv6 subnet calculator, or read the conventions behind the most common prefix on the /64 reference page.
Frequently asked questions
Is IPv6 faster than IPv4?
Not inherently. On well-run networks the throughput difference is negligible. IPv6 can feel slightly snappier on connections that escape carrier-grade NAT, because packets no longer get translated at the ISP, but routing quality and distance matter far more than protocol version.
Can IPv4 and IPv6 talk to each other directly?
No — they are separate protocols, and an IPv4-only host cannot open a session with an IPv6-only server by itself. Coexistence relies on dual-stack hosts that run both, tunnels that carry one inside the other, or translators such as NAT64 paired with DNS64.
Whatever happened to IPv5?
Version 5 was reserved in the late 1970s for an experimental streaming protocol called the Internet Stream Protocol (ST), which was designed to carry voice and video over the same addressing model as IPv4. It was never widely deployed, but its claim on version number 5 is why the next full revision became IPv6.
Do home routers support IPv6?
Nearly every consumer router sold in the last decade does, and Windows, macOS, Linux, Android and iOS have shipped it enabled by default for years. Whether you actually get v6 depends on your ISP: check the router's WAN status page for a global unicast address starting with 2xxx: or 3xxx:.
Should I disable IPv6 to fix network problems?
Rarely. Disabling it breaks features that prefer v6 (direct connectivity, some peer discovery) and masks the real fault. Modern stacks treat v4 and v6 in parallel using Happy Eyeballs, so a broken path usually fails over in milliseconds. Diagnose the actual problem instead of turning the protocol off.
Related tools and references
- IPv6 subnet calculator — expand, compress and analyze any IPv6 prefix.
- IPv6 /64 reference — why the standard LAN subnet holds 264 addresses.
- Subnet cheat sheet — all IPv4 prefixes from /0 to /32, printable.
- CIDR notation explained — the slash-prefix syntax both protocol versions share.