The great networking correction: frames reaffirmed

The paper considers how frame-based network technologies of local area networks are retaking their role as the primary campus technology, providing connections from desktop through backbone. This development comes at the expense of cell-based ATM, which was widely viewed as the inevitable, and highly desirable, next wave in transport technologies. ATM had been expected to supplant LAN-based transport and allow data, voice, and video to converge on a single, multi-transmission rate network. That ATM will become the dominant campus transport technology, however, seems less and less likely. Technical benefits once unique to ATM have become fewer with recent advances in both Ethernet and token ring technologies. As a result, monetary investment in ATM technology has fallen off sharply. ATM technology will likely become less visible as fewer servers and high-end workstations use it for direct connections. It will, of course, continue to provide robust switch-to-switch connections in some campus backbones. It may thrive in the carrier space, where its scalability, support for multiple traffic types, and circuit-based paradigm can be put to good use