More than a larger IPv4 address
IPv6 expands IPv4’s 32-bit addressing to 128 bits. Its base header rearranges fields and uses extension headers for additional information, so the formats differ beyond address length.
Upper layers and network devices must recognize the relevant version. IPv4 support alone does not establish that an application’s address handling, interfaces or deployment network support IPv6.
The fixed IPv6 header is 40 bytes. Payload Length excludes those bytes but includes following extension headers, and Next Header identifies the immediately following header rather than necessarily TCP or UDP. Forwarders decrement Hop Limit to constrain circulation; destination-node handling of zero is separately specified. There is no IPv4-style checksum in the base header, which does not remove transport checksums or every other error-detection mechanism.
Addresses and local neighbors
Addresses are commonly written as colon-separated hexadecimal groups with defined zero-compression rules. Unicast, multicast and anycast describe different delivery models; address text is neither a domain name nor a permanent identity.
Neighbor Discovery handles local link-layer address discovery, router discovery and neighbor reachability. IPv6 does not simply reuse IPv4 ARP packets for those tasks.
Fragmentation and protection boundaries
IPv6 fragmentation is performed by source nodes, not transit routers. Sources must account for the path’s supported packet size and relevant control messages.
Datagrams can still be lost or reordered. Larger addresses do not themselves encrypt or authenticate traffic; reliability, peer identity and application protection need their appropriate mechanisms.