IPv6 Subnet Calculator
Full 128-bit IPv6 subnetting, RFC 5952 compression/expansion, prefix ranges, and /64 or /48 block sizing.
Understanding IPv6 Address Architecture & Prefix Sizing
IPv6 expands the address space from 32 bits to 128 bits, representing addresses as eight 16-bit blocks (hextets) separated by colons. Unlike IPv4 where subnet masks are variable and broadcast addresses must be explicitly planned, IPv6 does away with broadcast entirely—replacing it with efficient multicast groups (such as ff02::1 for all nodes on a link).
In standard enterprise allocations:
- /48 Prefix: Assigned to entire enterprise campuses or multi-datacenter environments. Contains 65,536 /64 subnets.
- /56 Prefix: Assigned by residential and commercial ISPs to small-business customers. Contains 256 /64 subnets.
- /64 Prefix: The universal end-user subnet boundary required for SLAAC (Stateless Address Autoconfiguration). Provides 18,446,744,073,709,551,616 addresses.
- /127 Prefix (RFC 6164): Used specifically for inter-router point-to-point links to prevent ping-pong denial-of-service loops.
Related Network Utilities
Frequently Asked Questions
Why is a /64 prefix considered the standard subnet size in IPv6?
In IPv6 architecture (RFC 4291), a standard unicast address consists of a 64-bit routing prefix and a 64-bit Interface Identifier (IID). Stateless Address Autoconfiguration (SLAAC) strictly mandates a /64 prefix length to generate host addresses automatically from MAC addresses (EUI-64) or cryptographically random tokens (RFC 7217). Configuring prefixes longer than /64 (such as /80 or /112) breaks SLAAC and router advertisements.
What are the canonical compression rules in RFC 5952?
RFC 5952 establishes strict rules for consistent representation: 1) Leading zeros within a 16-bit hextet must be suppressed (e.g. 002a becomes 2a). 2) The double-colon :: must be used to shorten the longest run of consecutive 16-bit all-zero fields. 3) If there are multiple equal-length runs of zeros, the first run must be compressed. 4) The :: must not be used to shorten just a single 16-bit zero hextet. 5) Hexadecimal characters must always be lowercase.
What is the typical allocation size for an enterprise site in IPv6?
Regional Internet Registries (RIRs such as APNIC) and enterprise ISPs typically assign a /48 prefix to an enterprise organization. A single /48 block contains 65,536 distinct /64 subnets, each providing 18.4 quintillion (2^64) IP addresses—ample room for multi-VLAN segmentation, datacenter zones, and future branch expansions without network address translation (NAT).
Need this designed for a real production environment?
Consult with Gurinder Singh and vetted cloud specialists for Azure Landing Zones, AWS VPC peering, and hybrid connectivity.