What was the name of the first packet-switching node used in the ARPANET?

The first packet-switching node used in the ARPANET was the Interface Message Processor, usually abbreviated IMP.

The IMP was a specialized minicomputer that connected an ARPANET host computer to the network and handled the movement of packets between sites. Instead of requiring every host computer to manage networking directly, the design placed a dedicated communications computer between each host and the leased transmission lines.

Bolt, Beranek and Newman built the original IMPs in 1969 using ruggedized Honeywell DDP-516 minicomputers. The first IMP was installed at UCLA, where it connected to Leonard Kleinrock’s Network Measurement Center. A second node at the Stanford Research Institute soon established the first ARPANET host-to-host connection.

“Gateway,” “bridge,” and “router” can seem plausible because an IMP performed related networking functions. In fact, IMPs are often described as the first generation of gateways and as early ancestors of modern routers. However, the historical name for the ARPANET’s packet-switching node was Interface Message Processor, not one of those broader device categories.