Generated by GPT-5-mini| Ethernet II | |
|---|---|
| Name | Ethernet II |
| Caption | Schematic of an Ethernet II frame |
| Introduced | 1982 |
| Designer | Xerox PARC; DIX consortium |
| Standard | DIX Ethernet (1982) |
| Type | Data link layer protocol |
| Layer | Data Link Layer |
| Predecessor | Original Ethernet |
| Successor | IEEE 802.3 |
Ethernet II is the dominant frame format for wired LAN Ethernet deployments and a de facto standard for protocol multiplexing on Ethernet. It was formalized by the DEC, Intel, and Xerox (DIX) consortium and widely adopted in the early 1980s, enabling interoperable transmission of higher‑level protocols such as Internet Protocol and Address Resolution Protocol. Ethernet II remains prevalent across devices from Cisco Systems routers to Intel Corporation network interface controllers.
Ethernet II emerged from early work at Xerox PARC and subsequent commercialization by Xerox, Digital Equipment Corporation, and Intel culminating in the 1982 DIX specification, which followed experimental systems like Experimental Ethernet and implementations at PARC. The DIX specification coexisted with parallel standardization efforts at the IEEE, whose IEEE 802.3 project produced an alternate frame definition; disputes over framing and protocol identification led to real‑world interoperability testing among vendors such as DEC, IBM, HP, 3Com, and Sun Microsystems. The popularity of Internet Protocol stacks in UNIX systems and deployment by organizations like MIT and Bell Labs accelerated adoption, while industry consortia including the IETF and the World Wide Web Consortium later relied on Ethernet II framing for transport of IPv4 and IPv6. Over time, enhancements such as VLAN tagging (from IEEE 802.1Q) and EtherChannel aggregation (introduced by Cisco Systems) extended capabilities without replacing the basic Ethernet II frame semantics.
The Ethernet II frame consists of a sequence of fields starting with a 6‑octet destination MAC address and a 6‑octet source MAC address followed by a 2‑octet EtherType field, a payload (minimum 46, maximum 1500 octets), and a 4‑octet frame check sequence (FCS) using CRC‑32. Equipment vendors such as Intel Corporation, Realtek, and Broadcom implement this structure in hardware and firmware for NICs and switches used in products from Dell Technologies, Hewlett Packard Enterprise, and Lenovo. The frame layout interoperates with physical layer standards defined by IEEE 802.3 family members including IEEE 802.3u (Fast Ethernet), IEEE 802.3ab (Gigabit Ethernet over copper), and IEEE 802.3ae (10 Gigabit Ethernet), enabling link technologies like Twisted Pair cabling standardized by TIA/EIA and optical fiber transceivers from suppliers such as Finisar.
The 2‑octet EtherType field in the Ethernet II frame identifies the payload protocol (for example, 0x0800 for IPv4, 0x86DD for IPv6, 0x0806 for ARP). EtherType assignments are managed historically by the EtherType registry maintained by the IANA and were influenced by early registries and vendor allocations involving DEC, Xerox, and Intel. EtherType enables independent protocol stacks like IPsec, ARP, RARP, Reverse DNS services, and packet types used by AppleTalk and Novell NetWare. Operating systems such as Linux, FreeBSD, Windows NT, macOS, and Solaris use EtherType values in their network stack implementations to demultiplex frames to correct protocol handlers, while tools like tcpdump, Wireshark, and Netcat display EtherType to aid debugging and analysis.
Ethernet II is distinguished from IEEE 802.3 frames by its EtherType field; IEEE 802.3 originally used a length field and relied on an Logical Link Control header for protocol identification. Devices and stacks from Microsoft and AT&T historically supported both Ethernet II and IEEE 802.3/LLC to interoperate across heterogeneous networks. Modern switches and bridges from vendors including Juniper Networks and Arista Networks typically accept both frame types and rely on EtherType values or LLC SNAP headers (defined in IEEE 802.2 SNAP) to determine payload handling. Comparison points include frame field semantics, handling of small payloads (padding), and interaction with extensions like IEEE 802.1Q VLAN tags; real‑world networks often mix frames yet maintain compatibility via NIC firmware and OS drivers.
Ethernet II is implemented in hardware MAC controllers by companies such as Broadcom, Marvell Technology Group, Intel Corporation, and Realtek Semiconductor and is supported by operating system network stacks including Linux kernel, FreeBSD, NetBSD, OpenBSD, Microsoft Windows, and macOS. Enterprise deployments by organizations like Google, Amazon Web Services, Facebook (Meta Platforms), and Microsoft Azure use Ethernet II framing in data centers and backbone networks alongside technologies like VXLAN, MPLS, and BGP for transport and routing. Network appliances from Cisco Systems, Palo Alto Networks, and Fortinet process Ethernet II frames in switching, firewalling, and intrusion detection systems, often accelerated using DPDK or SmartNIC offload features provided by vendors like NVIDIA and Mellanox Technologies.
Ethernet II frames are commonly encapsulated or extended using mechanisms such as IEEE 802.1Q VLAN tagging, IEEE 802.1ad (QinQ) stacking, VXLAN tunneling, GRE, and IPsec transport modes. Carrier and metropolitan networks employ Pseudo‑Wire and MPLS label switching to carry Ethernet II payloads across wide area infrastructures standardized by organizations including the IETF and the Metro Ethernet Forum. Encapsulation protocols interact with EtherType semantics through SNAP headers or dedicated Ethertype values; products from Ciena, Huawei Technologies, Nokia, and Ericsson implement these combinations for carrier Ethernet services. As speeds increase with standards like IEEE 802.3by (25 Gigabit), IEEE 802.3bs (200/400 Gigabit), and emerging Terabit Ethernet research at IEEE 802.3, Ethernet II framing principles continue to underpin packet identification and delivery across networks spanning data centers, campuses, and carrier infrastructures.